understanding the factors that influence seedling establishment … · understanding the factors...
TRANSCRIPT
Understanding the factors that influence
seedling establishment to improve the
success of threatened species translocations
in the Mediterranean-climate region of
southwest Australia
By Christine Allen
Thesis presented for the degree of Doctor of Philosophy
The University of Western Australia
2014
Preface
i
Abstract
Translocations have become a common management tool across the globe to establish
or augment populations of threatened plant species to reduce the risk of extinction.
Translocations often involve germinating seed ex-situ and planting seedlings into a
specified site. As the translocation process is relatively costly and there is often limited
seed availability, it is important to maximise seedling survival to increase the chances
of establishing a self-sustaining population. Manipulating abiotic and biotic variables at
a translocation site is one strategy used to improve seedling survival and further
understand the ecology of threatened species. Southwest Australia is one of 34 global
biodiversity hotspots with a high proportion of threatened plant species.
Consequently, translocations have been implemented in this region by the Department
of Parks and Wildlife over the last 20 years with varied degrees of success. My research
focused on the Acacia and Banksia genera to build on the knowledge gained through
past translocations and tests some novel ideas to explore the questions: (1) Is it
possible to manipulate microhabitat “safe site” variables and plant seedlings with
particular characteristics that together improve their growth and survival in a
Mediterranean-climate? (2) Do plant physiological responses and microhabitat
measurements help to explain seedling survival and growth? (3) Can we use data on
natural seedling establishment and physiological responses to guide conservation of
threatened species in a drying climate? My thesis incorporates research from both
natural and translocation settings at three spatial scales.
Firstly, seedling establishment was investigated at the microhabitat scale with the
establishment of translocations of two Critically Endangered species Acacia
awestoniana (Chapter 2) and Banksia ionthocarpa subsp. ionthocarpa (Chapter 3) in
the Stirling Range region of southwest Australia. Experimental treatments were
incorporated at the level of individual plants (i.e. seedling age, planting season and
pre-translocation substrate) and at the microhabitat level (i.e. water supplementation
and location in regard to surrounding vegetation). Monitoring was undertaken for two
years. Monitoring of growth and survival of planted seedlings was complemented with
Preface
ii
measures of seedling health and environmental conditions. Seedling size at planting
was the most important variable explaining differences in survival of both translocated
species. Furthermore, planting seedlings of both species with a “plug” of potting mix
rather than bare roots improved seedling establishment and survival. Water
supplementation on a weekly basis generally improved growth of seedlings compared
to monthly watering, probably because the increase in soil moisture was short-lived at
both sites (i.e., less than seven days after watering). Microhabitat differences also
influenced the growth of both species, as the size of B. ionthocarpa seedlings was
negatively affected when surrounded by dense vegetation and thick litter. Acacia
awestoniana seedlings were significantly smaller when planted under a canopy species
with no water supplementation. Both translocations had a high overall survival rate
after two years which will assist with the conservation of these species through the
development of self-sustaining populations. Implementing an experimental approach
at seedling and microhabitat levels within the translocations has provided empirical
evidence for seedling establishment preferences which can greatly assist in future
translocations.
An understanding of seedling establishment in natural systems has the potential to
further improve seedling survival in a translocation setting. Therefore, Chapter 4 uses
surveys to investigate seedling establishment of common Acacia and Banksia species
in three post-fire sites. Survival and growth responses in the first summer after
recruitment were correlated with a range of microhabitat variables. Responses were
generally species-specific, with herbivory being the dominant factor influencing
survival. There was evidence of preferential herbivory of Acacia over Banksia
seedlings. Some other environmental variables that were associated with higher
seedling survival and growth included microtopographic depressions and ash layers.
Results of this study could be incorporated into future translocations such as planting
seedlings in artificial ash layers or under dead branches to reduce the chance of
herbivory.
Finally, Chapter 5 examines the adaptations of species and communities to drought at
a regional scale along an aridity gradient in southwest Australia including several
Preface
iii
Banksia and Acacia species. This study aimed to correlate an important drought
adaptation trait, leaf osmotic potential, with a range of climatic, microhabitat and leaf
functional traits in mature vegetation as well as seedlings. Leaf osmotic potential was
highly variable amongst species within a site, and community-averaged values did not
correlate well with measures of local water availability. However, as expected, species
at the driest site did have more negative osmotic potentials than those at the wettest
site. The osmotic potential of seedlings was generally less negative than their mature
counterparts. This may be due to higher soil moisture in the post-fire microhabitats, or
the inability of seedlings to express low values as it would lead to increased cell
expansion and thus larger evaporative losses. This study highlights the complexities of
incorporating physiological measures to identify differences in drought tolerance
across plant communities and provides some recommendations for future studies.
Overall, my thesis uses ecological research to improve management outcomes for the
translocation of threatened flora. This research is relevant in southwest Australia and
other Mediterranean-climate regions where land managers are responsible for
conserving an increasing proportion of threatened species in a drying climate.
Preface
iv
Table of Contents
Abstract .......................................................................................................................................... i
Table of Contents ......................................................................................................................... iv
Thesis Declaration ........................................................................................................................ vi
Acknowledgements ......................................................................................................................vii
Chapter 1: General Introduction ................................................................................................... 1
Abstract ..................................................................................................................................... 1
Threatened flora translocations ................................................................................................ 1
Seedling establishment in translocations and the “safe site” hypothesis ................................ 2
Seedling establishment in southwest Australia and other Mediterranean-climate regions .... 3
Adaptations to drought in Mediterranean-climate regions ...................................................... 5
Flora of southwest Australia ...................................................................................................... 8
Challenges to flora conservation in southwest Australia .......................................................... 9
Threatened flora translocations in southwest Australia ......................................................... 10
Thesis outline ........................................................................................................................... 14
Chapter 2: Incorporating a multi-level experimental approach in a threatened plant
translocation in the Mediterranean-climate region of southwest Australia .............................. 17
Abstract ................................................................................................................................... 17
introduction ............................................................................................................................. 18
Methods .................................................................................................................................. 22
Results ..................................................................................................................................... 32
Discussion ................................................................................................................................ 42
Chapter 3: Seedling size and microhabitat influence survival and growth of the Critically
Endangered Banksia ionthocarpa subsp. ionthocarpa in an experimental translocation .......... 49
Abstract ................................................................................................................................... 49
Introduction ............................................................................................................................. 50
Methods .................................................................................................................................. 54
Results ..................................................................................................................................... 62
Discussion ................................................................................................................................ 72
Chapter 4: The study of post-fire seedling establishment can guide decisions for conservation
management in Mediterranean-climate communities ............................................................... 80
Abstract ................................................................................................................................... 80
Introduction ............................................................................................................................. 80
Methods .................................................................................................................................. 84
Preface
v
Results ...................................................................................................................................... 89
Discussion................................................................................................................................. 96
Chapter 5: Using leaf osmotic potential to predict drought tolerance among vegetation
communities across an aridity gradient in southwest Australia ................................................ 103
Preface ................................................................................................................................... 103
Abstract .................................................................................................................................. 103
Introduction ........................................................................................................................... 104
Methods ................................................................................................................................. 108
Results .................................................................................................................................... 113
Discussion............................................................................................................................... 121
Chapter 6: General Discussion ................................................................................................... 127
Thesis aims ............................................................................................................................. 127
Recommendations ................................................................................................................. 133
Future research directions ..................................................................................................... 134
Conclusions ............................................................................................................................ 136
References ................................................................................................................................. 138
Appendices ................................................................................................................................. 156
Preface
vi
Thesis Declaration
This PhD thesis was completed while enrolled in the School of Plant Biology at the
University of Western Australia and has not been previously accepted for a degree at
this or any other institution.
The work contained in this thesis is my own, completed under the supervision of P
Poot, R Standish, M Moody and D Coates. Supervisors assisted with the experimental
design of all chapters and helped establish the field sites for Chapters 2, 3 and 4.
Supervisors also provided feedback on the chapter drafts. I collected the monitoring
data and conducted the statistical analyses for all chapters.
Preface
vii
Acknowledgements
And each day there is a chance that the mystery will surprise us
Or that we will surprise the mystery
An excerpt from ‘Days’ by Keren Gila Raiter, UWA Plant Biology PhD student
On the first day of my environmental science class in year 11, my teacher, Mr. Diegutis,
exclaimed “Isn’t life amazing!” and since then I have been hooked on learning about
biology. So firstly thank you to Mr. Diegutis, for igniting that spark of interest.
I first heard about the biodiversity hotspot in southwest Australia during my
undergraduate degree at the University of Wollongong. I was intrigued; it sounded like
an oasis of every type of plant imaginable (like a Wonka chocolate factory for plant
ecologists) and vowed that I would work there one day.
And here I am! What a journey of mysteries the last four years has been! I have been
privileged to work in some of the most beautiful places in southwest Australia to
recover the very species that make this region unique. Plus I have worked with some
equally beautiful people along the way. What more can a PhD student ask for!
A huge thanks to my supervisors for taking me on board and supporting me
throughout the rollercoaster ride that is a PhD. Pieter and Michael, I appreciate your
hard work before I had arrived in Perth to start the translocation approval process plus
the logistics of setting up two large scale translocations. Rachel, I appreciate our
thought-provoking discussions, plus your fresh ideas and encouraging words where
needed. Those wet and wild days in the field while tagging seedlings in Albany will
long be remembered! Dave, thank you for sharing your knowledge about conservation
in southwest Australia and helping me find my way through the bureaucracy during my
work with DPAW. And to all of you, your tireless work throughout all stages of my
PhD, particularly the editing of my chapters, is much appreciated. I really admire how
all of you can juggle so many pressures in the field of academia. This experience has
provided me with an insight into life as a scientist, and that is something I will draw
upon for many years to come!
I would like to acknowledge my main sources of funding which include the Australian
Postgraduate Award and Top-up scholarship from the University of Western Australia,
the Department of Parks and Wildlife external scholarship, plus a grant from the ANZ
Holsworth Wildlife Endowment. Furthermore, I appreciate receiving the Graduate
Preface
viii
Research School Travel Award and the Plant Biology Travel Grant which were used for
travel to my first international restoration conference in Merida, Mexico in 2011 and
the Ecological Society of Australia conference in Melbourne in 2012.
I have really enjoyed being part of the diverse group of Plant Biology staff and students
(I was often the token Australian!). The regular social events, including Friday morning
tea, Monday seminars, Ecofizz and Rottnest Summer School, make the School of Plant
Biology a special place to study and work. Thank you to Megan Ryan for your support
and encouragement. Special thanks to Barbara Jamieson, Pandy du Preez and Nat
Jagals, and all the other administration staff who keep the School running on a daily
basis. Thanks to the tireless work of the glasshouse staff Bill Piasini and Rob Creasy.
Additionally, thanks to Elizabeth Halladin and Greg Cawthray for their help with lab
equipment.
For the past four years, I have also been a part of the Ecosystem Restoration and
Intervention Ecology research group at UWA. Weekly meetings with this lively group
were often a highlight of my week. I have thoroughly enjoyed sharing the progress of
my own research and discussing the work of others in this friendly environment where
everyone is treated like family! So thank you to Richard Hobbs for leading this group
with wise words and lots of laughs. Plus all the members from past and present,
notably Bec Parson, Tim Morald, Kris Hulvey, Mike Perring, Mandy Trueman, Cristina
Ramalho, Lisa Denmead, Bridget Johnson, Keren Raiter, Mike Craig, Jo Burgar, Jodi
Price, Leonie Valentine, Melinda Moir, Maggie Triska, Hilary Harrop, Juan Garibello
and Mike Wysong.
My PhD would not have been possible without the support of the staff at DPAW in
Perth and Albany. Firstly, Bec Dillon, we could not have established the translocations
without your expertise and advice. Your good humour and smiling face during those
long days in the field made the manual labour a little easier! Leonie Monks, thank you
for your time and efforts to provide me with background information and advice about
the translocation program in southwest Australia. I am also very thankful to other
DPAW staff: Andrew Crawford, Sarah Barrett, Anne Cochrane, Michael Hislop and Bob
Dixon at Kings Park for their assistance with various parts of my project. Your
combined knowledge and experience in flora conservation in Western Australia is
remarkable!
Throughout the many days and weeks in the field I had the pleasure of getting to know
many volunteers. Volunteers were an integral aspect of all of my research chapters
and I am indebted to every one of them! A huge thank you to Bec Parsons, Nancy
Shackelford, Casey Causley, Merise Hocking, Amy Robinson, Maggie Triska (extra
thanks for the trailer driving!), Dylan Lehmann, Sonja Jakob, Anthea Challis, Emmaline
Preface
ix
Yearsley, Bridget Johnson, Keren Raiter, Jo Burgar, Nina Christiansen, Felipe Albornoz,
Kimberley Solly, Nicole Dakin, Marina Borges Osorio and Sarah Grain.
I have also been lucky enough to meet some people on this PhD journey who were my
Perth ‘family’ and have become great friends. Even though many are international
students, I know we will keep in touch throughout the years. Nancy Shackelford, with
patience and persistence you taught me how to drive a manual car and use the
statistical program ‘R’. These were two skills that were essential for completing my
PhD. We also had many laughs and thoughtful conversations along the way! Sonja
Jakob, the best office mate I could ask for! Thank you for being there in the good and
bad times. Jo Burgar, thank you for joining me on all those hippy outings (which often
ended in discussions about our work) and help with statistics and maps. Maggie Triska
(and adopted Allen), I appreciate your humour and obliging ways, I knew I could always
rely on you for help. Dr. Dini, your ever-positive attitude was annoying but infectious. I
think we just about solved the world’s problems during all of our morning coffees!
Annisa and Caroline Snowball, thank you for the fun holidays and activities that took
our minds off the PhD work (particularly while attempting to ice skate!)
And last but not least, my family and friends in Sydney. Thank you to my friends from
school and my undergraduate degree for your unwavering support and jokes about
‘Perff.’ To Mum, Dad, Em and Liz, you have been the backbone of my support network.
You were always there to pick me up when I fell down. All of your words of
encouragement and care were with me every step of the PhD. I really couldn’t have
asked for a more loving family.
I dedicate my PhD to the diversity of flora that is often overlooked by decision makers
of today.
May we, as a society, learn to appreciate the natural environment for all of its mystery.
Chapter 1
1
Chapter 1
General Introduction
ABSTRACT
With flora extinctions occurring across the globe at an ever-increasing rate, land managers
are using a range of measures to prevent further species losses. Traditionally these measures
include in-situ conservation through the establishment of protected areas and protection of
flora with legislation. These approaches continue to be effective but as populations become
smaller and more isolated, direct intervention is now considered an important option to
reduce the risk of extinction. Intervention can frequently include creating or augmenting
populations through translocations of seedlings germinated ex-situ. Southwest Australia has
a high proportion of threatened species and translocations have been implemented in this
region by the Department of Parks and Wildlife over the last 20 years with varied degrees of
success. My research builds on the knowledge gained through past translocations and tests
some novel ideas, in the context of an increasingly dry climate, to explore the questions: (1)
Is it possible to manipulate microhabitat “safe site” variables and plant seedlings with
particular characteristics that together improve their growth and survival in a
Mediterranean-climate? (2) Do plant physiological responses and microhabitat
measurements help to explain seedling survival and growth? (3) Can we use data on natural
seedling establishment and physiological responses to guide conservation of threatened
species in a drying climate?
THREATENED FLORA TRANSLOCATIONS
Translocations are the deliberate introduction of a plant (or animal) species to a site
where it may or may not have occurred historically (IUCN, 1998). Translocations have
become an important method to protect rare and threatened species from extinction
around the globe. Translocations are a much debated practice as they are resource
intensive and have had varying degrees of success (Guerrant, 2013). Success in itself is
often difficult to define, and will differ for long and short-term goals (Guerrant & Kaye,
2007). Pavlik (1996) outlined four main attributes to help define translocation goals
and ultimately translocation success for plants:
Chapter 1
2
1. Abundance: ability for a translocated population to establish, grow and reach
reproductive maturity (short-term goal)
2. Extent: ability of the translocated population to expand beyond the area of
translocation with the presence of dispersal agents (short-term goal)
3. Resilience: the capacity for the translocated population to survive stochastic
events such as fire (long-term goal)
4. Persistence: the translocated population can be self-sustaining and functions
(such as pollination by native pollinators) within an ecological community (long-
term goal)
It has been argued that plant translocations should be regarded as experiments with
multiple treatments, to improve our understanding of the factors that determine
translocation success (Allen, 1994; Gordon, 1996; Guerrant & Kaye, 2007; Sarrazin &
Barbault, 1996). The development of a translocation management plan requires an in-
depth understanding of the target species’ biology, ecology and physiological
tolerances. Given the variability in species’ requirements, it is difficult to establish one
“best practice” guideline for plant translocations (Morgan, 1999). However, some
methodologies almost invariably lead to higher overall success. For example,
translocations of seedlings often have higher rates of survival compared with those
that rely on the application of broadcast seeds (Dalrymple et al., 2012). High survival at
these early establishment stages is an integral aspect of achieving the first goal of
translocations and has thus been subject to considerable research in the past.
SEEDLING ESTABLISHMENT IN TRANSLOCATIONS AND THE “SAFE
SITE” HYPOTHESIS
Germinating seeds ex-situ and planting seedlings to establish a translocation, has
become a common protocol to conserve threatened plant species across the globe
(Dalrymple et al., 2012). However, the seedling establishment stage has been
recognised as a “bottleneck” of a plant lifecycle in many species due to the low survival
Chapter 1
3
rate in the first year after germination (sometimes 0.1–1% survival in first year)
(Arrieta & Suárez, 2006; Gómez-Aparicio, 2008). Harper (1977) coined the term “safe
sites” to describe sites with the ideal conditions for seeds to germinate and survive to
reproductive maturity. The safe site hypothesis assumes a lottery-type system where a
pool of seeds land in highly heterogeneous microhabitats. Subsequently a species’
morphological and physiological features will determine if the seed can germinate and
recruit in a specific microhabitat. Urbanska (1997) extended the use of this concept to
the restoration of disturbed habitats by identifying the characteristics and availability
of safe sites in reference (undisturbed) ecosystems. These safe sites could then be
artificially created within a disturbed site to maximise survival of propagules. In this
way, the safe site hypothesis could also be further explored in the conservation of
threatened species. Rare and threatened species can sometimes have highly
specialised safe site requirements and failed translocations have been attributed to a
lack of knowledge of these requirements (Morgan, 1999; Primack, 1996). Therefore,
incorporating the safe site hypothesis into a translocation setting using seedlings
rather than seeds can provide a theoretical framework within which microhabitats can
be manipulated to create safe sites. In practice, using this framework may involve
planting translocated seedlings in a range of microhabitats and monitoring their
subsequent growth and survival. Concurrent measurements of environmental factors
within each microhabitat would provide important insights into which factors are
driving differences in growth and survival.
SEEDLING ESTABLISHMENT IN SOUTHWEST AUSTRALIA AND OTHER
MEDITERRANEAN-CLIMATE REGIONS
The species investigated in this thesis are endemic to southwest Australia which
typically has a Mediterranean climate. Mediterranean-climate regions are found in five
locations across the globe (Mediterranean Basin, California, central Chile, the Cape
Region in South Africa and southwest Australia). These regions are characterised by
dry and hot summers with wet, mild winters and the vegetation is dominated by
Chapter 1
4
evergreen sclerophyllous species (Cowling et al., 1996). With only three to four months
of winter rainfall, there is a small window of opportunity for seedling germination and
recruitment during winter and spring. Observational surveys and experimental
manipulations have been used in an attempt to define a safe site in Mediterranean-
climate regions (Castro et al., 2005; Gómez-Aparicio et al., 2005; Ibáñez & Schupp,
2002; Lloret et al., 2005; Petru & Menges, 2003). These studies demonstrated that
drought is indeed a constraining factor for seedling establishment. Therefore, abiotic
and biotic factors that ameliorate water stress such as shading, litter and
supplementing water were shown to be important elements of a safe site.
Alongside drought, fire is a main disturbance in Mediterranean-climate regions and a
major driver of recruitment events, as evidenced by the flush of plant recruitment in
the months following a fire event (Carrington & Keeley, 1999). Woody species from
these regions usually respond to fire along a spectrum of the two following extremes
(1) resprouters survive fire and regenerate with heat resistant structures on
lignotubers or trunks, whereas (2) reseeders are killed by fire and regenerate solely by
seed (Bell, 2001; Pate et al., 1990). Therefore, the safe site hypothesis is relevant to
reseeding species as seeds land in a heterogeneous post-fire landscape. Successful
germination and establishment will depend on species-specific characteristics and
tolerances to that specific microhabitat. Post-fire environments are often
characterised by fertile ash and low levels of adult plant competition, both providing
ideal conditions for seedling establishment (Buhk et al., 2007). However, these
environments are also very exposed (i.e. high solar radiation and wind speeds) which
facilitate high levels of herbivory, and hydrophobic ash can reduce soil water
absorption (Bradstock, 1991; Tyler, 1995). Understanding the conditions under which
post-fire seedling recruitment occurs in a Mediterranean-climate environment can give
clues as to microhabitat conditions that could be manipulated to improve
translocation success.
Chapter 1
5
ADAPTATIONS TO DROUGHT IN MEDITERRANEAN-CLIMATE
REGIONS
Most woody plants growing in Mediterranean-climate regions will face water deficits
regularly throughout their life cycle. Therefore, morphological and physiological
adaptations have evolved to optimise plant function under low-water availability.
These adaptations involve either avoiding or tolerating low soil water potential by
maintaining a positive gradient between the soil and the leaves (i.e. leaves need to be
‘drier’ than the soil) to avoid desiccation (Figure 1). Some morphological adaptations
to avoid desiccation include, reductions in leaf area to reduce water demand
(Clemente et al., 2005; Valladares & Sánchez-Gómez, 2006) and increased allocation to
roots to increase water supply (Hernández et al., 2010; Otieno et al., 2001). While
morphological traits are often less plastic in response to environmental stressors such
as drought, physiological responses are usually plastic and rapid (Table 1; Pohlman et
al., 2005; Valladares & Sánchez-Gómez, 2006) Most species use a combination of
physiological and morphological adaptations to cope with drought which may vary
with age.
Figure 1 – General adaptations to survive drought include avoidance or tolerance. Adapted
from Kozlowski & Pallardy (2002).
Chapter 1
6
Drought tolerance in woody plant seedlings
The time between post-fire germination and the onset of summer drought can vary
considerably from year to year. Therefore, seedlings have to develop desiccation
tolerance/ avoidance soon after germination. Some species avoid desiccation by
predominantly allocating resources to root growth shortly after germination to tap
into deeper and more reliable water sources (Enright & Lamont, 1992; Kozlowski &
Pallardy, 2002; Padilla et al., 2007; Padilla & Pugnaire, 2007). However, at the onset of
summer drought, the roots of many seedlings are still relatively shallow and prone to
desiccation. Consequently, seedlings also use physiological mechanisms to limit water
loss such as a reduction in stomatal conductance and an increase in leaf solute
concentration i.e. osmotic adjustment (Vilagrosa et al., 2003; Warren et al., 2007). The
latter allows seedlings to maintain turgor under drier conditions. Seedlings can use
these physiological mechanisms to a point, but if low water availability in the shallow
soil is prolonged, irreversible damage can occur in cells. In particular, hydraulic failure
due to xylem embolism is known to be one of the main causes of seedling mortality in
the field (McDowell et al., 2008; Mediavilla & Escudero, 2004; Williams et al., 1997).
Avoiding desiccation with gas exchange
Drought stress directly affects gas exchange in plants by reducing the amount of soil
water available for metabolic functions and for transpiration. Whereas some species
reduce leaf stomatal conductance to prevent leaf water potentials from falling
(‘isohydric’), others allow leaf water potentials to fall (‘anisohydric’) in order to
generate a larger driving force for water transport from the soil (McDowell et al.,
2008). While mature plants can be highly plastic in these responses due to previous
exposure to environmental conditions, seedlings are often less able to control their
water status (Clemente et al., 2005). Therefore, early detection of stress is often
desirable when planting seedlings for translocation and/or restoration programs.
Measuring stomatal conductance, chlorophyll fluorescence and water potential can all
detect temperature and water stress in leaves (Kooten & Snel, 1990; Turner, 1981).
Chapter 1
7
However, it is often difficult or impractical to apply these methods to young seedlings
due to the small size of their leaves and their vulnerability. Therefore, less invasive
methods that monitor seedling physiological activity would be preferable. One such
method is infrared thermography which can determine the surface temperature of
leaves. When combined with ambient air temperature, this method can provide an
estimate of leaf stomatal conductance (Jones, 2004). While infrared thermography has
been used widely in agriculture (Cohen et al., 2005; Jones et al., 2002; Moller et al.,
2007; Oerke et al., 2006), its application in native ecosystems has been limited. This
method is promising for application on seedlings in a restoration or conservation
setting, particularly when working with rare and threatened species when
conventional methods, such as harvesting whole seedlings, is not always desired.
Tolerating desiccation with changes in leaf osmotic potential
Concentrating solutes in leaves is a strategy used by many woody plants to avoid
dehydration (Figure 1; Kozlowski & Pallardy, 2002; Yazaki et al., 2010). Some species
maintain a high leaf solute concentration (i.e. low osmotic potential) throughout the
year. This generates lower leaf water potentials and increases the driving force for
water transport from the soil to the leaves. It also enables them to maintain turgor
(cell rigidity) at lower leaf water potentials. Other species use this mechanism only
during the build up to the dry season by actively concentrating solutes in their leaf cells
in response to drying soil, a process known as osmotic adjustment (Morgan, 2003).
Measurements of osmotic adjustment have been widely used as an indicator of
drought tolerance of species or cultivars in agriculture (Babu et al., 1999; Lilley &
Ludlow, 1996; Morgan, 1992; Morgan, 1984). Additionally, a meta-analysis by Bartlett
et al. (2012) found that osmotic potential at turgor loss point (wilting) was closely
correlated to local water availability in plant species from a range of biomes,
demonstrating the importance of using osmotic potential to tolerate drought.
However, there is little known about detecting differences in osmotic potential across
a smaller regional scale, particularly in woody seedlings in a field environment. These
Chapter 1
8
measures of drought tolerance could become increasingly important when considering
translocations in a drying climate.
Integrating physiological measurements into flora conservation programs is still in its
infancy but can complement traditional monitoring such as survival and growth
measures to identify underlying causality in patterns of growth and survival (Carey,
2005; Cooke et al., 2013). For example, physiological measurements can be used to
identify favourable microhabitats where seedlings have access to higher soil water as
well as predicting seedling mortality based on stress levels. Furthermore, Monks et al.
(2012) suggested that physiological health (i.e. transpiration and photosynthetic
efficiency) of translocated individuals could be compared with reference populations
to determine if stress is inhibiting growth or maturity in the former. This information
could be useful for predicting if a translocated population eventually becoming self-
sustaining.
FLORA OF SOUTHWEST AUSTRALIA
Southwest Australia is a Mediterranean fire-prone region with old, climatically
buffered and infertile landscapes (Hopper, 2009). It has been suggested that the
unique combination of a relatively stable climate and long geological history has led to
a mosaic of different soil types to which plant species have adapted (Hopper & Gioia,
2004). The diversity of soil types over relatively short distances may have led to an
evolutionary advantage for those seeds and seedlings that remained close to the
maternal plant. The low dispersal rates observed in flora throughout southwest
Australia may thus be a reflection of the region’s evolutionary history (Hopper, 2009).
Reduced dispersal often leads to increased genetic divergence of local populations
which can promote speciation and ultimately the generation of local endemics
(Hopper, 2009; Hopper & Gioia, 2004). As a consequence, southwest Australia is
characterised by a large proportion of narrowly distributed local endemics.
Approximately 50% of the 7380 described plant species in southwest Australia are
Chapter 1
9
endemic to the region, with many populations having less than 100 individuals
(Cowling et al., 1996; Hopper & Gioia, 2004).
CHALLENGES TO FLORA CONSERVATION IN SOUTHWEST
AUSTRALIA
Southwest Australia is one of 34 global biodiversity hotspots due to its high number of
endemic plant species combined with widespread human related disturbance
(Mittermeier et al., 2011). Approximately 70% of native vegetation has been removed
since European settlement. Other major threats to plant species in the region include
disease, invasion by introduced species, dryland salinity and altered fire regimes. As a
consequence, many populations of naturally rare species have become locally extinct
and further fragmented and are now confined to small patches of remnant vegetation
(Hopper & Gioia, 2004). Recent alteration of fire regimes by humans has modified the
frequency of regenerations in fire-adapted communities (Bell, 2001). Inappropriate fire
regimes have led to immature or senescing plant communities. Senescing communities
are often characterised by no available resources or space for seedling establishment.
In addition, seed viability has been shown to substantially decrease in some species
due to long intervals between fires which further limits the recruitment potential of
many reseeder species. For example, Verticordia fimbrilepis ssp. fimbrilepis is an
endangered, reseeding species that is found in small populations and recruits
prolifically after fire. This species is likely to rely on fire for population persistence and
local extinction was predicted at sites where fire had been excluded (Yates & Ladd,
2010).
In Western Australia, the Department of Parks and Wildlife (DPAW) has responsibility
for the conservation of the native flora throughout the state. DPAW uses a ranking
system based on the IUCN Red List for Threatened Plants criteria for prioritising and
managing threatened species (Coates & Atkins, 2001). The categories are based on the
level of threat in order to prioritise conservation actions (Table 2). Recovery plans are
developed for the Threatened Flora, termed Declared Rare Flora (DRF) under the State
Chapter 1
10
Wildlife Conservation Act 1953. These recovery plans outline strategies to protect or
reduce harmful impacts on existing populations. Developing recovery plans for the
Critically Endangered species is often a priority. For those species that are on the brink
of extinction, establishing new populations or restocking existing populations by
translocations has become an important conservation tool.
Table 2 – Threatened Flora (Declared Rare Flora) ranking used by the Department of Parks and
Wildlife and based on the IUCN Red List Criteria (adapted from Smith et al. 2013).
Threatened Flora Ranking Significance
CR: Critically endangered Extremely high risk of extinction in the wild
EN: Endangered Very high risk of extinction in the wild
VU: Vulnerable High risk of extinction in the wild
THREATENED FLORA TRANSLOCATIONS IN SOUTHWEST AUSTRALIA
Threatened flora translocations commenced in 1994 by The Department of Parks and
Wildlife, Western Australia and became part of a dedicated translocation program in
1998. Since then approximately 60 translocations have been implemented in 105
locations (Coates, D. pers. comm., January 2011). Over time, these translocations have
had varying degrees of success and protocols that have led to a higher survival, such as
fencing to reduce herbivory and germinating seeds ex-situ and planting seedlings, have
become standard practices (Jusaitis, 2005; Monks et al., 2012; Monks & Coates, 2002).
Other experimental manipulations employed at the local microhabitat scale include
shading, burning before seeding, watering, and mulching (Cochrane et al., 2000). Some
treatments, such as mulching, did not have the hypothesised effect of increasing
survival of Acacia aprica and Daviesia bursarioides (Guerrant, 2012). Currently, there
are no published syntheses about the outcomes of various experimental treatments
used in past translocations in southwest Australia. Furthermore, some treatments
Chapter 1
11
were applied to all planted seedlings and a control group was not used. Experimental
controls have rarely been incorporated in translocations across the globe, with
controls implemented in only ~15 % of 48 studies that had manipulated habitats at
pre-translocation stage and ~25 % of 65 studies at the post-translocation stage
(Guerrant, 2012). This highlights a substantial gap in the knowledge which could be
used to assist conservation managers such as Department of Parks and Wildlife in the
assessment of planting sites. As a consequence, there is a need to establish
translocations in a systematic manner with a range of treatments at different stages of
the translocation i.e. manipulating microhabitats before planting and using different
pre-translocation growth conditions for seedlings. Using this multi-level method has
the potential to provide a detailed insight into local optimal conditions for seedling
establishment which could consequently improve translocation success.
Using translocations to conserve Acacia and Banksia species
Acacia and Banksia are iconic Australian genera that are widespread across the
continent, and both genera, particularly in Western Australia, contain a high
proportion of species threatened with extinction. Southwest Australia accounts for
80% of Banksia diversity in Australia and these species often dominate sandplain
vegetation communities (Lamont & Connell, 1996). Fragmentation is a major threat to
Banksia species and given the high level of fragmentation of the southwest region
there are twenty one threatened Banksia species across Western Australia (Table 3;
Lamont et al., 2007; Smith, 2013). Southwest Australia is one of three primary centres
in Australia for Acacia species richness and have a much wider distribution compared
with Banksia species including more inland lower rainfall zones (González-Orozco et
al., 2011). The species richness of Acacia is highest in Western Australia’s wheatbelt
which has been highly fragmented (~95% cleared) for agriculture (Buist et al., 2002;
Hopper & Maslin, 1978). Consequently, there are currently thirty three threatened
Acacia species in Western Australia, most of which occur throughout the wheatbelt
region (Table 3).
Chapter 1
12
Table 3 – The total number of described and Threatened (Declared Rare Flora) Banksia and
Acacia species in Western Australia (as of September 2013) based on their level of threat using
the IUCN Red List criteria and ranked by the Department of Parks and Wildlife (Chapman &
Gioia, 2013; Department of Parks and Wildlife, 2013; Smith, 2013).
Genus
# Described species
#
Vulnerable
#
Endangered
# Critically
Endangered
Total number of Threatened species
Banksia 158
8
5
7 20
Acacia 624 11 7 15 33
To date, a total of fourteen threatened Banksia and Acacia species (seven for each
genus, including the translocations reported in this thesis) have been translocated by
the Department of Parks and Wildlife (Monks, L pers. comm., November 2013). These
translocations have included a range of microhabitat treatments, most of which did
not have an effect on seedling survival after two years (Table 4). Watering and fencing
were applied to all translocated seedlings. While the effectiveness of fencing varied
between translocated species (Table 4; A. aprica vs. A. cochlocarpa subsp.
cochlocarpa), the effects of watering frequency have not been tested by the
Department thus far. Survival of seedlings was ~80–90% after two years for ten of the
twelve studies included (Table 4). A notable exception is the translocation of Banksia
ionthocarpa (which is the focus of Chapter 3). Furthermore, the survival of seedlings
often varied widely between different planting batches. The reasons for differences in
mortality of seedlings after planting was sometimes attributed to drought but other
reasons were largely unknown. It is thought that an understanding of the biology and
ecology of both genera might assist with choosing safe sites for seedlings to establish
successfully.
13
Chapter 1
Translocated Species Treatments Survival after two years
Banksia anatona
(two sites)
Soil conditioner (TerraCottem®) applied to half of seedlings.
All plants watered and fenced.
Site 1 = 71–98 %
Site 2 = 84–95% No effect of soil conditioner
Banksia brownii (three translocation sites
established for three different genetic forms)
Soil conditioner (TerraCottem®) applied to half of seedlings (only southern and
northern form translocations). All plants watered and fenced.
98–100% for all translocations.
No effect of soil conditioner
Banksia ionthocarpa Shading with shade cloth applied to half of seedlings 23%
No effect of shading
Banksia montana (two sites)
No treatments (all seedlings watered and fenced) Site 1= 89–100% Site 2 = 100%
Banksia nivea subsp. uliginosa Planted seeds
Heavy vs. Light top soil
88%
No effect of topsoil
Banksia oreophila No phosphite vs. Phosphite. 93% survival (1 year)
Acacia aprica Mulching around half of seedlings
Fenced vs. Unfenced
88.7% mulched vs. 87.5% unmulched.
62% fenced vs. 38% unfenced (1 year).
Acacia cochlocarpa subsp. cochlocarpa Seedling age (9 months vs. 18 months)
Watered vs. Unwatered and Fenced vs. Unfenced
88–96%
No effect of treatments
Acacia imitans No treatments (all plants watered and fenced) 94%
Acacia subflexuosa subsp. capillata No treatments (all plants watered and fenced) First planting = 100% (4 years)
Second planting = 53% survival (3 years)
Acacia unguicula No treatments (all plants watered and fenced) 88 %
Acacia volubilis No treatments (seedlings fenced and watered in first and second planting,
fencing only in third)
First planting = 13%; Second planting = 59%;
Third Planting = 0%
Table 4 – Summary of past translocations of Acacia and Banksia species conducted by the Department of Parks and Wildlife (the translocation of B. nivea subsp. uliginosa
was established by Kings Park Botanic Gardens). All translocations were introduction types, apart from A. volubilis that was an augmentation type. Information for the
translocation of Banksia cuneata was unavailable. Survival rates after two years or otherwise stated in brackets (Sourced from DPAW unpublished data and Dixon, B pers.
comm. March, 2014; Department of Conservation and Land Management, 2004; Monks, 1999; Monks & Coates, 2002).
Chapter 1
14
Species in both genera establish predominately after fire with a range of regeneration
strategies from facultative resprouting to reseeding. However, Banksia and Acacia
species are likely to differ in their safe site requirements due to differences in
dispersal, post-fire behaviour and nutrient acquisition. Acacia are regarded as
disturbance opportunists due to their effective nitrogen fixation strategy and
subsequent rapid establishment after fire (Adams et al., 2010). Banksia species are
often bradysporous (retain their seeds within their canopy for a long period of time)
and usually have larger seeds that do not disperse far from the parent plant and
produce large seedlings after fire. Banksia seedlings can rapidly produce tap roots up
to two metres deep in their first year to survive drought conditions over summer
(Lamont & Groom, 1998). Therefore, it could be predicted that Banksia seedlings
prefer safe sites that are close to parent plants (seeds drop under burnt plants after
fire) with less compacted soils so the roots can reach deeper layers. In contrast, it is
predicted that safe sites for Acacia seedlings would be more varied as seeds can be
dispersed more widely by ants and birds. Acacia seedlings are also likely to survive in
more exposed microhabitats which are a feature of recently burnt sites.
THESIS OUTLINE
Anthropogenic changes continue to threaten plant species across the globe, and there
is an urgent need to develop efficient strategies to conserve biodiversity and prevent
extinctions (Pressey et al., 2007). Translocations are one method of active
management that has had some success in creating or augmenting populations of
threatened species. However, translocations are often costly and resource-intensive,
face seed supply limitations and are especially hampered by poor knowledge of a
species’ ecology and life history. Therefore, establishing experimental translocations
with seedlings planted in a range of local microhabitats can greatly improve our
understanding of a species’ ecology as well as maximising the chance of at least some
survival and growth. Within a Mediterranean context, water availability is likely one of
the main limiting factors influencing seedling survival and growth. Therefore, it is
Chapter 1
15
intuitive that increasing soil moisture content using direct or indirect methods is likely
to improve seedling survival. Water supplementation and competitor removal are both
methods that can achieve this goal. Although water supplementation has been used
successfully in earlier translocations in southwest Australia, it is still unclear about
what watering frequencies and quantities would optimise seedling survival over the
first few summers. This thesis focuses on the Acacia and Banksia genera and explores
the association between water availability and seedling establishment within natural
and translocation settings at three spatial scales.
Chapter 2 and 3 focus on seedling establishment at the microhabitat scale through
experimental translocations of two Critically Endangered species in southwest
Australia: Acacia awestoniana (Chapter 2) and Banksia ionthocarpa (Chapter 3). A
multi-level experiment was used for both species in which treatments were assigned at
the seedling and the plot level. Habitat characteristics of the last remaining
populations for these species were used to inform choice of microhabitat treatments.
In addition, some more unconventional treatments were tested such as clearing of
vegetation before planting to remove competition, and planting in different seasons.
Seedling growth and survival were monitored for two years and were complemented
with measures of environmental conditions and seedling health.
Chapter 4 tests the concept of safe sites in natural settings at a landscape scale. This
study aimed to identify the dominant environmental factors influencing survival and
growth of seedlings of a range of Acacia and Banksia species in their first summer of
establishment. Microhabitat variables were surveyed around newly emerged seedlings
at three post-fire sites. It was hypothesised that microhabitats with a higher water-
holding capacity would increase first-summer survival and thus would be an important
characteristic of safe sites. Results of this study were envisioned to inform how to
further improve site manipulations to optimise seedling survival in future
translocations in southwest Australia.
Chapter 5 was intended to focus in more detail on the specific drought adaptations
that A. awestoniana (Chapter 2) employs in comparison with a range of other common
Chapter 1
16
as well as rare Acacia species in a glasshouse study. However, as explained in the
preface to Chapter 5, a fungal disease caused such high mortality that this experiment
had to be aborted. Instead, Chapter 5 now investigates adaptations to drought for
species and communities at a regional scale along an aridity gradient in southwest
Australia including several Banksia and Acacia species. As there is relatively little
known about plant community water relations in this region, this study aimed to
obtain baseline data for variation in leaf osmotic potential and a range of other
potentially correlated variables amongst the dominant woody species (seedlings as
well as mature vegetation) in post-fire regeneration sites. It was predicted that plants
in wetter regions would have relatively less negative osmotic potentials in November
2012 (pre-summer) and April 2013 (post-summer) compared with those located in
drier locations. This study contributes to the knowledge of drought adaptations of
native species and can potentially indicate which species would be most at risk under
future hotter and drier climate scenarios.
By incorporating ecological theories and practice to guide the management of
threatened species, my research is pertinent at a global scale given the ongoing threat
of species loss due to habitat clearance and fragmentation as well as climate change
(Fahrig, 2003; Giam et al., 2010; Malcolm et al., 2006). Understanding the main factors
that drive seedling establishment and survival will assist with the development of
translocation protocols for threatened species.
Chapter 2
17
Chapter 2
Incorporating a multi-level experimental approach
in a threatened plant translocation in the
Mediterranean-climate region of southwest
Australia
ABSTRACT
Experimental translocations are used to reduce the risk of flora extinctions but so far have
had varied levels of success. Planting seedlings in different microhabitats, as well as
manipulating microhabitat factors and pre-translocation conditions can assist in
understanding local seedling requirements which can contribute to developing protocols to
improve future translocation success. This study aimed to significantly increase one of the
two remaining populations of the Critically Endangered Acacia awestoniana in southwest
Australia. At the same time, experimental manipulations were incorporated at the pre-
translocation and microhabitat levels to enhance our ecological understanding of
establishment requirements. The survival and height of planted seedlings was recorded for
two years. These measurements were complemented with infrared leaf temperature and
stomatal conductance measurements during summer as an estimate of seedling drought
stress. After two years, seedlings planted at six months old had significantly higher survival
rates than those planted at four months old (90 versus 70%). They also were more likely to
produce flowering buds before the end of the monitoring period, despite suffering higher
levels of herbivory. Although watering and microhabitat plot treatments did not significantly
influence seedling survival, there were large effects on plant growth. Seedlings in plots with
weekly water supplementation grew significantly (~60%) taller and had lower leaf
temperatures indicative of less drought stress, compared with seedlings in the no watering
treatments. Planting seedlings in different microhabitats had a significant effect on seedling
growth but this was dependent on the age of the seedlings at planting. This result is likely
due to the interaction between plant size at establishment and a range of both facilitative
and competitive effects. Based on these results a number or recommendations are made
which are considered likely to improve future translocation success of threatened plant
species in seasonally dry climates.
Chapter 2
18
INTRODUCTION
Threatened species translocations are increasingly being implemented across the
globe to reduce the risk of extinction. However, augmenting existing populations or
establishing new ones is often a resource intensive process and the outcomes are
often uncertain. To better understand the reasons for failure or success of flora
translocations, several researchers have emphasised the need to undertake more
experimental manipulations (Allen, 1994; Gordon, 1996; Guerrant & Kaye, 2007;
Sarrazin & Barbault, 1996). Past translocation studies have manipulated levels of
competition (Burmeier & Jensen, 2009; Gordon, 1996; Pavlik et al., 1993), herbivory
(Jusaitis, 2005; Monks & Coates, 2002), litter (Holl & Hayes, 2006) and light intensity
(Peterson et al., 2013; Rowland & Maun, 2001). In a review of global plant
translocations, Godefroid et al. (2011) found that incorporating pre or post-planting
management (such as removing competition, watering or soil loosening) into the
translocation process, generally had a positive effect on seedling survival in the first
year of establishment. Overall, experimental translocations can improve the
mechanistic understanding of the abiotic and biotic factors which influence seedling
establishment.
Planting seedlings often leads to higher overall survival rates within translocations
when compared to direct seeding (Dalrymple et al., 2012). However, this method also
has several disadvantages. Planting seedlings is costly and resource intensive and initial
mortality is usually quite high (Guerrant, 2013; Monks et al., 2012). This result has
been ascribed to possible root damage inflicted during transplanting and to initially
poor root-soil contact (Kozlowski & Pallardy, 2002). Furthermore, seedlings can
become root bound in their pre-translocation containers which may hamper root
growth when planted at the translocation site. This could subsequently compromise
their water uptake capacity during the dry season (Green et al., 2005). Growing
seedlings in sand and planting them with bare roots rather than with a plug of potting
mix has the potential to improve root growth in field soil as seedlings will be forced to
grow directly into the field soil. Indeed, manipulating the pre-translocation substrate
can potentially increase survival after transplantation, yet this important aspect of a
Chapter 2
19
seedling’s growth and development is not often included in experimental
translocations (Guerrant, 2012; Maunder, 1992)
Seedling age at planting has the potential to influence survival and growth as well.
Planting older and thus larger seedlings in the field usually improves survival and
growth which has been attributed to a larger root: shoot volume. This makes seedlings
more resilient to drought and competition with other vegetation (Cuesta et al., 2010).
However, Verona (2012) found that seedlings of three species of woody shrub from
Spain were more drought stressed after maturing in the nursery for three years
instead of one year before planting. The three year old seedlings had greater shoot to
root ratios, and consequently experienced greater drought stress in the field with
significant leaf death compared to the younger seedlings. Therefore, testing the
optimal age and size of seedlings by varying the time in nursery conditions has the
potential to further increase the chance of survival and growth.
Summer drought is one of the main causes of seedling mortality in Mediterranean-
climate regions such as southwest Australia (Castro et al., 2005; Garrido et al., 2007;
Hnatiuk & Hopkins, 1980). Thus, supplementing water by irrigation is commonly used
in translocations. Although the benefits of watering seem obvious, it is a costly
strategy, and the frequency and volume of supplemented water optimal for survival
and growth is often unknown. Apart from irrigation, water availability can also be
enhanced through indirect methods. For example, the location of plantings in respect
to surrounding vegetation can also affect the amount of water available to
translocated seedlings (Ibáñez & Schupp, 2001). These effects can be positive
(reducing evaporative demand by shading, hydraulic lift of water from deeper soil
layers) as well as negative (competition for water and nutrients). The effect of
surrounding vegetation may depend on factors such as transplanting season, soil
texture and local climate (Ibáñez & Schupp, 2001; Lloret et al., 2005). Experimental
translocations designed to investigate these interactions can potentially increase the
chance of future translocation success.
Monitoring translocated plants is imperative to determine ultimate translocation
success (Allen, 1994; Menges, 2008; Sheean et al., 2012). In many past translocations,
Chapter 2
20
monitoring has focused on growth and survival following planting and most
translocations have not considered the physiological responses of seedlings.
Physiological activity, when expressed as photosynthetic rate or stomatal conductance,
can estimate drought stress and may be a good predictor of future plant mortality.
These measurements could be useful in plant conservation research, yet equipment is
not always widely available and translocated seedlings are often too small and
vulnerable to be measured (Cooke & Suski, 2008). However, non-invasive techniques
such as thermal imaging can be used as a proxy for leaf transpiration levels and thus
plant water stress in the field (Jones, 1999).
Southwest Australia is one of 35 global biodiversity hotspots with approximately 3,000
endemic plant species, many of which are narrowly distributed and poorly known
(Mittermeier et al., 2011). With widespread anthropogenic fragmentation and
disturbance throughout this region, many endemic species are currently on the verge
of extinction. Translocations have been employed to increase population numbers and
sizes of threatened species, and most translocations have targeted species which are
currently classified as Critically Endangered. Some sixty plant species have been the
subject of translocations by the Western Australia Department of Parks and Wildlife
since 1998. The Department routinely collects seeds of threatened plant species,
germinates them ex-situ and then uses the seedlings to establish populations on new
sites or augment the original populations (Cochrane et al., 2007). Seventy five percent
of past translocations in southwest Australia have supplemented water (often with
automated irrigation systems), during the first two summers after planting (Monks, L
pers. comm. 29th Nov 2013). Supplementing water has generally improved the survival
of these translocated seedlings, however, Monks & Coates (2002) found that watered
and non-watered Acacia cochlocarpa subsp. cochlocarpa seedlings had similar survival
rates (~90%) two years after translocation. These high survival rates were attributed
to maturity of seedlings when planted (9 and 18 months). Therefore, manipulating age
at planting and watering frequency within a translocation can disentangle the relative
importance of these two factors in the early establishment stages. This can lead to the
most efficient use of resources while most likely increasing the numbers of
translocated seedlings that reach reproductive maturity.
Chapter 2
21
Acacia awestoniana is a threatened species listed under the Western Australian
Wildlife Conservation Act of 1953 and under the Commonwealth EPBC Act 2000. It is
ranked as Critically Endangered and is confined to two small populations in the Stirling
Range National Park, Western Australia. Acacia awestoniana has shown a rapid decline
in population size, distribution and general plant health since 1996 (Department of
Environment and Conservation, 2009). This decline is probably related to a
combination of natural senescence and poor recruitment after fire. The present study
focuses on an experimental translocation established for A. awestoniana seedlings in
2010. The main aims of this translocation were to (1) augment the smaller of the two
remaining populations to reduce small population effects such as in-breeding and loss
of genetic diversity (Ellstrand & Elam, 1993) and generate a self-sustaining population,
and (2) to obtain empirical evidence regarding the factors that drive translocation
success. A multi-level experimental approach was used to test the effects of pre-
translocation treatments i.e. substrate (sand or potting mix) and seedling age, and plot
treatments i.e. watering frequency and microhabitat (beneath overstorey trees or in
the open), on seedling survival and growth. I hypothesised that frequently watered,
older seedlings, which were grown in sand and positioned in the gaps between
overstorey trees would have the highest survival and growth rates. The latter was
inferred from earlier investigations by Lamont (1985) who suggested that the low
density of understorey shrubs in these Eucalyptus wandoo woodlands was due to the
competitive suppression by shallow roots of E. wandoo. Over the first two years,
traditional monitoring measurements (i.e. survival and growth) were complemented
by infrared thermography of seedlings and porometry to obtain a better mechanistic
understanding of differences in seedling survival and growth.
Chapter 2
22
METHODS
Study species
Acacia awestoniana is a spreading perennial shrub to three metres tall with resinous
branches and oval phyllodes. It is an obligate seeder that is killed by fire and relies on
a soil stored seed bank for recruitment (Department of Environment and Conservation,
2009). Acacia awestoniana flowers between September and November and insects are
suspected to be the main pollinators (Department of Environment and Conservation,
2008). In 1996, there were approximately 2,300 mature plants found in two
populations. Since then, there has been a severe decline in adult numbers likely due to
natural senescence in one population, while a fire killed 98% of mature plants in the
second population in 2006. Although there was substantial seedling recruitment after
the fire, many seedlings were heavily grazed by rabbits and /or kangaroos (Department
of Environment and Conservation, 2009). As a result, the two remaining populations
now occur in an area of less than three hectares and contain a total of 77 mature
plants and 337 juveniles (Barrett, S pers. comm. 12th July 2013) (Table 1). Given the
current small size of these populations, reduced genetic diversity and inbreeding
depression are legitimate concerns.
Table 1 – Numbers of A. awestoniana adult and juveniles in the two remaining populations.
Population two is made up of three sub - populations (a, b and c). Both populations are found
within the Stirling Range National Park. Coordinates are not provided due to the conservation
status of the species.
Year Population Mature Individuals Juvenile Individuals
2011 1* 6 1 2012 2a 13 122 2011 2b 50 200 2011 2c 8 14
* site of this experimental translocation.
Germination
The A. awestoniana seeds used for the current study were collected from population
two (due to the low number of seeds available from population one) by the
Department of Parks and Wildlife in 1996 and 2009. Seeds were germinated from
twenty one maternal lines. There was no clear differentiation between survival and
Chapter 2
23
growth among seedlings from different maternal lines and the results are not reported
in this study. To facilitate germination, arils were removed from the seeds and the
seed coat was nicked to overcome physical dormancy and then plated onto 7.5 g/l agar
in Petri dishes. All seeds were germinated in a controlled temperature room at 15 °C
with a 12 hour photoperiod.
We tested two seedling level treatments (Table 2). In order to test the effect of
seedling age, batches of seed were set to germinate on two separate occasions: 15th
January 2010 (batch A) and 8th April 2010 (batch B). After germination, the seedlings
were transferred to small cell trays and moved to the Kings Park and Botanic Gardens
nursery. When the seedlings started producing roots (2–3 days), they were transferred
to standard 50 mm × 50 mm seedling tubes filled with one of two substrates. Batch A
was planted between 12–15th July 2010 (ca. 6 months old), while batch B was planted
between 21–22nd August 2010 (ca. 4 months old). Evidently, possible differences
between the two batches can be due to seedling age as well as planting date.
Hereafter, this treatment will be referred to as seedling age.
Seedling substrate
We tested the effect of pre-translocation substrate in the experiment. Seedlings were
either planted with soil intact (hereafter ‘potting-mix seedlings’) which is the standard
practice for translocations in southwest Australia, or with roots directly exposed to site
soil (hereafter ‘sand seedlings’). The potting mix substrate consisted of jarrah sawdust,
river sand and coarse sand (ratio 2:1:1/2). A fertilizer mixture was added to the potting
mix including Osmocote slow release native fertiliser (low phosphorous plus trace
elements 3 kg/m3), ferrous sulphate (500 g/m3), dolomite (2.3 kg/m3) and lime (2.5
kg/m3). The sand substrate consisted of washed river sand and slow release fertiliser (3
kg/m3). The seedlings were placed in glasshouse conditions for 2–4 weeks and then
allowed to acclimatise outdoors for 4–6 months (autumn – winter).
Chapter 2
24
Experimental set-up
The translocation site was located just within the boundary of Stirling Range National
Park, Western Australia, and adjacent to population one (at approximately 150
metres). Population one was chosen to be augmented due to the small number of
remaining individuals (Table 1). The site is characterised by open Eucalyptus wandoo
woodland with associated sedge and grass species Loxocarya cinerea, Lepidosperma
brunonianum, and Neurachne alopecuroidea (Hocking, 2011). The soil consisted of a
sandy loam top soil over clay. Twenty four 3.5 × 3.5 m fenced plots were marked out
randomly within a 200 × 200 m area. Plot position was chosen to test the possible
facilitative or competitive effects of E. wandoo on the planted seedlings. Twelve plots
were positioned less than two metres from the base of a mature E. wandoo tree, while
another twelve plots were positioned in the open areas in between trees (at least four
metres from the closest E. wandoo).
Thirty six A. awestoniana seedlings were randomly allocated to each plot and planted
in a square grid (6 x 6 seedlings) with 30 cm between seedlings. In total 864 seedlings
were planted. Existing vegetation within the plots consisted of sparse grasses and
sedges and was left intact. Each plot had an equal representation of seedling level
treatments (i.e. pre-translocation substrate, seedling age and maternal line) (Table 2).
Coded metal tags were secured in the ground with stainless steel pegs for seedling
identification.
Soil moisture manipulation
To determine the effect of watering on plant growth and survival, a water tank with a
solar powered pump, developed by the Department of Parks and Wildlife for its plant
translocation program, was set up on-site in November 2010. The tank was filled on a
monthly basis with chlorinated local town water. Plots were randomly assigned one of
three treatments: 1) weekly automatic drip irrigation; 2) monthly hand-watering and 3)
no watering. The automatic drip irrigation treatment reflected protocol that has
become common practice in other threatened species translocations in southwest
Chapter 2
25
Australia. Similar to past translocations, one litre of water per seedling per week was
supplemented over the summer months (Dillon, R pers. comm. October 2010).
Drippers were set at a rate of 70 ml per minute and operated once a week before
dawn. The monthly hand watering treatment was chosen as a possible alternative to
setting up an expensive automated treatment. This method could potentially be useful
for translocations that are situated relatively close to human habitation (e.g. where
local land managers or other volunteers can assist with watering). This treatment also
would simulate a once per month summer storm event which is consistent with
weather patterns in this region. Once monthly over summer, 144 L of water was slowly
spread over each treatment plot to allow for soil infiltration. This volume was
calculated as one litre per week for each seedling in the plot. The remaining plots were
designated as control plots with no extra water supplementation. Irrigation treatments
operated from December to March for two consecutive summers (2010 – 2012).
Table 2 – Summary of experimental treatments at the plot and seedling level in the translocation of Acacia awestoniana
Experimental Treatment Treatment Levels
Plot treatments
Microhabitat 1. <2 m from E. wandoo tree base 2. >4 m from E. wandoo tree base
Water supplementation 1. Weekly (drip irrigation) 2. Monthly (manual) 3. None
Seedling treatments
Pre - planting substrate 1. Potting mix 2. Sand
Seedling age/ time of planting 1. Four months old (planted August 2010) 2. Six months old (planted July 2010)
Chapter 2
26
Soil moisture measurements
A neutron moisture probe (CPN 503DR-1.5, ICT International Pty Ltd) was used to
measure relative soil moisture content in every plot over the two year monitoring
period. Two metre deep holes were drilled in the centre of each plot (the closest
seedling was at least 15cm away) for soil moisture measurements (Track Mounted Drill
Rig, EVH Drill Engineering Pty Ltd). A mixture of bentonite clay and water was poured
into the holes and PVC tubes (~60 mm diameter and closed at the bottom) were then
inserted. Thereafter, pipes were shook to release any air bubbles from the clay mixture
which would interfere with the neutron probe readings. A rim of piping was left
exposed to hold the probe in place and capped to prevent water from entering. Soil
moisture was measured six times throughout the monitoring period (January, March
and June/ July) in the afternoons one month after the last manual water
supplementation and one week after the last irrigation (during summer
measurements). The soil profile within the small area of the translocation site was
assumed to be relatively homogeneous, and therefore raw neutron counts were used
as a measure of relative water availability across plots. Neutron counts were measured
after a 16 second period from a soil depth of 0 cm to 180 cm with 20 cm intervals. The
raw counts from each plot were converted to adjusted raw counts by subtracting
December 2010 raw counts (measured prior to first watering) of each specific plot with
raw counts after watering treatments began (Hanson & Dickey, 1993). This allowed all
post watering measurements to be adjusted to variations among plots before
watering. To simplify analysis, the adjusted raw counts from neutron moisture
measurements were divided into three soil depths: 20–60 cm, 60–100 cm and 100–
180 cm. The 0–20 cm neutron count was discarded as some neutrons can escape into
the air leading to inaccurate soil moisture readings close to the soil surface (Chanasyk
& Naeth, 1996).
Plant measurements
Seedling height was used as a measure of growth for A. awestoniana, as this species
has an erect growth habit. Nine monitoring trips were conducted between January
Chapter 2
27
2011 and July 2012 to record survival. During six of these trips, seedling height was
measured and visual estimates of the level of leaf herbivory (% seedlings with leaf
damage per plot) were conducted in June 2011 and July 2012. During the final
monitoring trip in July 2012, some plants had developed flower buds. The number of
budding plants was recorded in each plot and identities of the budding plants were
correlated with seedling and plot level treatments.
During March 2011 and 2012, leaf surface temperature of seedlings was determined
using an infrared camera as an estimate for plant transpiration and stomatal
conductance (TiR-32 Thermal Imager, Fluke Thermography, Washington, USA). Leaf
temperature is dependent on surface evaporative cooling through plant stomates.
Leaves with limited water supply generally lower their transpiration rates by reducing
their stomatal conductance which results in higher leaf temperatures. Thus, leaves are
often considerably warmer than ambient air temperature when a plant is under water
stress while a similar or lower temperature would indicate that water is readily
available (Jones, 1999). Five plants in twelve of the plots, representing all treatment
combinations, were randomly chosen for infrared thermography measurements.
Images were taken 24 hours after weekly and monthly watering was undertaken from
8 am to 11 am. Images of seedlings were taken randomly throughout this time to
reduce bias from increasing ambient temperatures. The weather conditions were
sunny with less than 10% cloud cover throughout this period. Images were taken of the
top of each plant and parallel to each target leaf. Directly comparing absolute
temperatures between separate infrared images can be ambiguous (Grant et al.,
2007). This is partly due to changes in local environmental conditions and to a slight
drift in mean camera calibration due to internal electronics heating (Jones et al., 2002).
Therefore, I used a custom-made aluminium plate (5 × 5 cm) to calibrate the camera
temperatures in each image separately. The plate was painted matt black and had a
type K thermocouple (response time 0.1 seconds) attached to the back. In each image,
the plate was positioned next to the target plant and held up by a metal pin that was
attached to the back (Figure 1). The plate’s thermocouple temperature reading was
compared with its infrared temperature reading, and the offset was used to correct
the leaf infrared temperatures in the picture. Simultaneously, ambient air temperature
Chapter 2
28
was recorded for each image with another type K thermocouple (response time 0.1
seconds). This thermocouple was housed in a vented PVC case and painted white to
reflect radiated heat and allow air to flow freely. Both thermocouples were connected
to AD597 thermocouple conditioners and the output was calculated in an analogue to
digital converter and displayed at the front of the casing via a microprocessor (Hortin,
P pers. comm., 8th September 2013) (Figure 1). Images were analysed with imaging
software provided with the camera (SmartView version 3.0.126.0, Fluke
Thermography, Washington, USA). The perimeters of approximately five leaves per
seedling were outlined and the software calculated their average temperature using
the data from pixels in the selected area.
To evaluate the relationship between leaf temperature and leaf stomatal conductance,
both were measured simultaneously during the March 2012 monitoring trip on a
cloudless and calm day. If leaf temperature is strongly correlated with stomatal
conductance then infrared thermography could be used as a relatively easy measure of
plant physiological response to drought. Stomatal conductance was measured using a
porometer (SC-1 Leaf Porometer, Decagon Devices Inc, Washington, USA) on the
youngest fully expanded phyllodes from seedlings in all plot treatment combinations.
Acacia phyllodes have stomates on both sides so stomatal conductance was measured
on both sides and an average value was calculated for use in subsequent analyses.
Chapter 2
29
Figure 1 – Equipment used for infrared leaf temperature measurements. A custom-made
black reference surface (foreground; with thermocouple soldered to the back), ambient air
temperature casing (white), and digital display device, all positioned around the target
seedling.
Statistical analyses
Linear mixed models within the nlme package in R (R Core Team, 2013) were used to
analyse the effects of seedling and plot level treatments on survival and height of A.
awestoniana seedlings. The maximal models contained all explanatory variables and
their interactions (Table 3). Due to the complexity of the experimental design, separate
models were created for each monitoring month rather than using a repeated
measures analysis. Plot was used as a random factor in all models. For the categorical
variables “watering”, “location to E. wandoo”, “pre-translocation substrate” and
“seedling age/ time of planting” the reference levels for each of these variables used in
the mixed models were no water, open, potting mix and batch A respectively. Soil
depth was a categorical variable with three levels used in the soil moisture analysis and
deviation coding was used which compared the mean soil moisture of each depth level
with the grand mean of all depths (UCLA: Statistical Consulting Group, 2013). Non-
significant variables were deleted from each model in a step-wise manner. A Maximum
Chapter 2
30
Likelihood (ML) approach was used during step-wise deletion to find parameters
closely correlated to the probability of the data (Bolker, Brooks et al., 2009). Maximal
and reduced models were compared with an ANOVA and the model with the lowest
Akaike Information Criteria (AIC) was judged to be the optimal model (Johnson &
Omland, 2004; Zuur, Ieno et al., 2007). Diagnostic plots were used to identify
heterogeneous variances by plotting residuals against each of the explanatory
variables. When heterogeneity was detected, slopes were allowed to vary with the
heterogeneous factor (Table 3). The optimal models were analysed using Restricted
Maximum Likelihood (REML) which incorporates the random effect and therefore final
predictions are less biased (Bolker et al., 2009).
Soil moisture measurements and leaf-air temperature differences were analysed in
separate models with plot treatments used as explanatory variables (Table 3). To
evaluate the general trend of soil moisture over time, an autoregressive variance-
covariance term was added as months since planting. This allowed correlation to
decrease with increased temporal distance between measurements (Zuur et al., 2007).
Time was also included in the main treatment effects in the soil moisture model
(Thaxton et al., 2012). When analysing leaf- air temperature difference, there was no
difference between proximity to E. wandoo so data from seedlings in both
microhabitats were pooled. Mixed models with soil moisture and leaf to air
temperature as response variables were reduced as above to find the optimal model. A
generalised linear model was used to compare stomatal conductance (measured from
top of phyllode) and leaf to air temperature to see if the two measures were
correlated.
Chapter 2
31
Table 3 – The response and treatment variables used in linear mixed models to analyse the
survival, growth, herbivory and drought-stress of translocated A. awestoniana seedlings, and
the effects of the plot-level experimental treatments on soil moisture. A binomial distribution
(survival=1, dead=0) was used to analyse survival and presence of herbivory, while a normal
distribution was used for all other models. Separate models were created for each monitoring
month.
Response variable Treatment variables included Heterogeneous
variances
Survival
and
Presence of leaf herbivory
1. Planting age 2. Pre-translocation soil
substrate 3. *Initial height of
seedlings used as a co-variable
4. Plot watering frequency 5. Plot proximity to E.
wandoo
Height (mm) 1. Planting age 2. Pre planting soil substrate 3. Plot watering frequency 4. Plot proximity to E.
wandoo
Slopes varied with
planting age
Soil moisture (adjusted count ratio)
1. Plot watering frequency 2. Plot proximity to E.
wandoo 3. Soil depth (20-60, 60-
100cm and 100-180cm) 4. Time (months since
planting)
Slopes varied with
soil depth
Leaf to air temperature
difference
1. Plot watering frequency
*(Metz, Sousa et al., 2010; Thaxton, Cordell et al., 2012)
Chapter 2
32
RESULTS
Climate over monitoring period
During the first year after translocation (July 2010 – Sept 2011), the North Stirling
climate station (22.8 km north of translocation site) measured 6–88% lower than
average rainfall in every month except for January 2011. January 2011 had a fourfold
increase in rainfall which was spread across the entire month (Figure 2) (Bureau of
Meteorology, 2013). The three month period between October and December 2011
was relatively wet, while below-average rainfall conditions returned in early 2012 until
May 2012.
Figure 2 – Observed monthly total rainfall (mm) at the Bureau of Meteorology North Stirling
climate station (22.8 km from translocation site) over the monitoring period compared to the
long term rainfall (1958-2010).
0
10
20
30
40
50
60
70
80
90
Apr-10 Jul-10 Oct-10 Jan-11 Apr-11 Jul-11 Oct-11 Jan-12 Apr-12 Jul-12
Rai
nfal
l (m
m)
Monitoring month
Observed monthly total Long term monthly total Batch A planting
(6 months old)
Batch B planting (4 months old)
First summer Second summer
Chapter 2
33
Soil moisture
Soil moisture was generally correlated with local rainfall. This was most apparent in
the upper 20–60 cm which showed significantly greater variation over the monitoring
period compared with the deeper layers (Figure 2 & 3, Appendix 1 : depth x time
interaction). The highest values were recorded in the first summer (January 2011) after
four times above average rainfall, and in July 2012 (mid-winter) at the end of the
monitoring period. Lower soil moisture as a consequence of no watering was only
evident in the intermediate soil layer (i.e. 60–100 cm; Appendix 1: watering treatment
x depth interaction). This pattern could be attributed to the timing of measurement, as
moisture contents were recorded just before watering events and thus soils could have
dried out for a week or a month (depending on watering treatment), especially over
summer. Proximity to E. wandoo did not have a consistent effect on moisture content,
although in the deepest soil layer, soil moisture was generally higher under the canopy
of E. wandoo (Appendix 1: microhabitat x depth interaction).
Chapter 2
34
Figure 3 – Time course of relative soil moisture content ± SE, as obtained by neutron moisture
probes, and expressed as the ratio of raw counts compared to Dec 2010 counts. The two plot
treatments which include watering frequency over summer (No, monthly and weekly
watering)and location to Eucalyptus wandoo (O = open and W = under E.wandoo) are included
between January 2011 and July 2012. A) Upper soil depth (20–60 cm), B) mid soil depth (60–
100 cm) and C) lower soil depth (100–180 cm).
Chapter 2
35
Seedling Survival
After two years of monitoring, the overall survival of planted A. awestoniana seedlings
was high at 81% (July 2012). The younger seedlings (batch B) had a significantly lower
probability of surviving compared to the older and taller seedlings (batch A) (Dec 2010
p=0.0001; July 2012 p<0.0001; Figure 4). Part of this difference in survival was
apparent before the onset of irrigation treatments (December 2010; batch A and B =
~98% and ~87% survival respectively) but was exacerbated in the December 2010 –
March 2011 period (Figure 4). At the seedling level, initial height was strongly
correlated with seedling age and had a significant effect on survival (Appendix 2).
Furthermore, within each batch, the smaller seedlings generally had a significantly
lower chance of survival (Table 4). There was also a significant interaction between
planting substrate and planting age, with seedlings grown in potting mix having a
higher survival than those grown in sand but only for the older seedlings of batch A
(Figure 4). By July 2012, batch A seedlings in potting mix were ~5% more likely to
survive than those in sand, while batch B seedlings had similar survival regardless of
pre-translocation substrate (Appendix 2).
Figure 4 – Average survival per plot ± SE (n = 24) of Acacia awestoniana seedlings throughout
the monitoring period based on the seedling level treatment seedling age (6 months old
planted in July 2010, versus 4 months old planted in August 2010), and planting substrate
(potting mix versus sand).
50 55 60 65 70 75 80 85 90 95
100
Dec-10 Mar-11 Jun-11 Sep-11 Dec-11 Mar-12 Jun-12
Perc
enta
ge s
urvi
val
Monitoring month
6 months/ potting 6 months/ sand 4 months/ potting 4 months/ sand
Chapter 2
36
Table 4 – Average initial height (mm) ± SE at planting (July/ August 2010) of seedlings that
survived (alive) or died (dead) during the two year monitoring period. Seedlings are
categorised based on their seedling age (Batch A: 6 months planted in July 2010 and batch B: 4
months planted in August 2010) and pre-translocation substrate (potting or sand). The
numbers of seedlings included in each category are included in brackets. Bold values indicate
significant height differences between alive and dead seedlings (p<0.05).
Seedling age Substrate Alive Dead p- value
Batch A: 6 months old Potting 163 ± 5 (207) 127 ± 22 (9) 0.10
planted in July Sand 265 ± 5 (191) 197 ± 19 (24) 0.002
Batch B: 4 months old Potting 34 ± 1 (135) 25 ± 1 (61) <0.001
planted in August Sand 35 ± 1 (121) 27 ± 1 (58) <0.001
There was no relationship between plot treatments (watering and location to E.
wandoo) and seedling survival throughout the monitoring period, although the plots
located in the open without watering tended to have the lowest survival (Table 5).
Table 5 – Average seedling survival per plot (%) ± SE (n = 4 plots per treatment combination),
two years after planting (July 2012) as related to plot treatments: watering frequency and
location to E. wandoo. There were no significant differences in seedling survival as related to
plot treatments.
Average % seedling survival
Watering treatments Plot location
Under E. wandoo Open
No water 83 ± 5 73 ± 6
Monthly 81 ± 8 82 ± 6
Weekly 83 ± 5 83 ± 4
Seedling Growth
Initial seedling height at planting depended on pre-translocation substrate and
seedling age, with batch A seedlings being substantially taller than batch B seedlings.
Furthermore, the height of batch A seedlings were further differentiated by substrate
as sand seedlings were significantly taller than potting mix seedlings (Figure 5,
Appendix 3, substrate x batch interaction). After planting, potting mix grown plants
Chapter 2
37
had a higher rate of growth than sand grown plants (Figure 5). However, after March
2011 (end of first summer), the relative height differences between seedlings grown
on the different substrates remained constant.
Figure 5 – Seedling height (mm) ± SE of translocated A. awestoniana seedlings per plot (n =
24) over two years based on their seedling age (6 months old in July 2010 and 4 months old in
August 2010) and pre-translocation substrate. The first point represents seedling height at
planting.
Weekly watering generally had a strong positive effect on seedling height which was
significant at the end of the first summer (Figure 6; Appendix 3). The monthly watering
and no watering treatments both resulted in similarly lower growth rates compared
with weekly watering. Location of plot in relation to E. wandoo did not have a
consistent effect on seedling height, and was partly dependent on seedling size at
planting (i.e. batch A versus batch B). While batch A seedlings were tallest in plots that
were watered weekly and located in the open, batch B seedlings showed greater
growth in plots placed under E. wandoo trees, especially in the first few months after
watering (Figure 6; three way interaction, Appendix 3). The smallest seedlings in both
batches tended to be located in plots under E. wandoo with no water
0
50
100
150
200
250
300
350
400
450
Aug-10 Nov-10 Feb-11 May-11 Aug-11 Nov-11 Feb-12 May-12
Seed
ling
heig
ht (m
m)
Monitoring month
6 months/ potting 6 months/ sand 4 months/ potting 4 months/ sand
Chapter 2
38
0
100
200
300
400
500
Aug-10 Nov-10 Feb-11 May-11 Aug-11 Nov-11 Feb-12 May-12
Seed
ling
heig
ht (m
m)
Monitoring month
Weekly-O
Monthly-O
No water-O
Weekly-W
Monthly-W
No water-W
supplementation. Note that some seedlings showed slight decreases in height over the
winter period which was due to herbivory (pers. obs.)
Figure 6 – Average seedling height (mm) per plot (n = 24) ± SE over monitoring time as
dependent on planting age (circles = 6 months old; crosses = 4 months old) and plot
treatments: dotted lines (W = under E. wandoo) versus solid lines (O = open) and watering
supplementation. The first point represents the initial height at planting (either July or August
2010). Water was supplemented only during the Jan – Mar periods.
Seedling herbivory and flowering
Levels of leaf herbivory, expressed as % of seedlings with herbivore damage per plot,
were significantly higher in plots located under E. wandoo canopies compared to those
in open plots (Figure 7; p <0.001). Furthermore, seedlings that were smaller at planting
(batch B) were significantly less likely to have evidence of herbivory compared to the
taller batch A seedlings (d.f. 568; p = 0.04 July 2011 and d.f. 542; p <0.001 July 2012).
Herbivory was higher in the first year after transplant and became more variable the
second (Figure 7). The watering treatments did not significantly affect the likelihood of
herbivore damage. Herbivores usually targeted new growth (pers. obs.) but there was
no relationship with presence of herbivory and seedling death. Personal observations
Chapter 2
39
during the monitoring period suggest that herbivory was especially due to caterpillars
and snails.
Figure 7 – Average percentage of seedlings per plot (number of seedlings in brackets) ± SE with
leaf herbivore damage as related to plot treatments: water supplementation (weekly, monthly
and no watering over summer) and plot location (O = open and W = under E. wandoo) in June
2011 and July 2012.
During the last monitoring trip in July 2012, 4% (n = 29) of seedlings had developed
flower buds. Budding seedlings were differentiated by planting age with 86% of
budding seedlings from batch A (p = 0.007). Seedlings with buds were significantly
taller (~620 ± 40 cm) than those without (~290 ± 6 cm; d.f. 29 p<0.001). There was no
relationship between budding and planting substrate or plot treatments.
Physiological measurements
Infrared thermography was used during the first and second summer of monitoring on
a selected number of seedlings as a proxy for leaf physiological activity. Twenty four
hours after watering, significant differences in leaf temperature between the watering
treatments were only observed during the second summer. Seedlings in weekly or
monthly watered plots had leaf temperatures similar to ambient, while seedlings with
0
10
20
30
40
50
60
70
80
90
100
Jun-11 Jul-12
Pla
nts
with
he
rbiv
ore
dam
age
(%)
Monitoring month
Weekly - O (n = 144)
Weekly - W (n = 141)
Monthly - O (n = 142)
Monthly - W (n = 144)
No Water - O (n = 144)
No Water - W (n = 143)
Chapter 2
40
no water supplementation had on average 2 °C higher than ambient leaf temperatures
(Figure 8, no watering vs. weekly watering d.f. 10 p = 0.03, no watering vs. monthly
watering d.f. 10 p = 0.05). The leaf-air temperature differences were significant though
weakly related to stomatal conductance as obtained by leaf porometry (Figure 9, p =
0.02). Leaves that were warmer than ambient air temperature had a lower stomatal
conductance.
Chapter 2
41
Figure 8 – The leaf- air temperature difference (°C) ± SE for selected seedlings as dependent on
plot watering treatments (n = 13 weekly, n = 11 monthly and n = 22 no water). Measurements
were obtained by a combination of infrared imaging and local ambient air temperature
measurements in March 2012. Images were taken from the top of A. awestoniana seedlings.
Figure 9 – Relationship between stomatal conductance (as obtained by porometry) and leaf-
temperature difference (as obtained with infrared thermography) of leaves from selected A.
awestoniana seedlings (n = 46). Measurements were taken simultaneously on the same plants
in March 2012. The line represents a significant linear regression (p = 0.02; R2 = 0.22).
-1
0
1
2
3
Weekly Monthly No Water
Leaf
- ai
r tem
pera
ture
di
ffere
nce
(°C
)
Watering treatment
R² = 0.22
-4 -3 -2 -1 0 1 2 3 4 5 6
0 50 100 150 200
Leaf
-air
tem
pera
ture
di
ffere
nce
(°C
)
Stomatal conductance (mmol m-2 s-1)
Chapter 2
42
DISCUSSION
The translocation of the Critically Endangered A. awestoniana during the early
establishment stages was quite successful. At the end of the two-year monitoring
period, 81% of the seedlings had survived with 4% of these having already entered the
reproductive phase. In addition, the multi-level experimental design has greatly
enhanced our understanding of the main factors driving translocation success of this
species.
Seedling survival
The age of seedlings at planting was the most important variable affecting survival. As
expected, planting age was closely related to initial seedling height, with the older
batch A seedlings approximately ten times taller than the batch B seedlings. However,
the batch B seedlings were also planted one month later and thus the observed
difference in survival could also be a timing effect. However, the tallest of seedlings of
both batches were more likely to survive after two years which suggests that seedling
size was the dominant factor. It is well-known that larger seedlings often have greater
survival and growth compared to smaller seedlings in both natural (Gilbert et al., 2001)
and planted systems (Cuesta et al., 2010). Large seedlings usually have longer root
systems with access to water in deeper layers which has been directly linked to higher
survival throughout summer drought (Cuesta et al., 2010; Padilla et al., 2007; Padilla &
Pugnaire, 2007). Also, larger seedlings are able to reach maturity in a shorter time
period. Indeed, buds were mainly observed on batch A seedlings at the end of the two
year monitoring period. Reaching reproductive maturity in a shorter time frame could
assist with achieving population persistence sooner which is a key factor in
determining ultimate translocation success (Guerrant & Kaye, 2007; Kaye, 2008).
While planting age and seedling maturity have a strong influence on survival after
planting, they have not been regularly included as experimental treatments in
threatened species translocations. Maschinski & Duquesnel (2007) translocated
different size classes of the Endangered palm Pseudophoenix sargentii, and found that
taller seedlings (> 0.83 m) had significantly higher survival than smaller transplants and
Chapter 2
43
had improved chances to reach reproductive maturity. Restoration studies have found
a general trend towards increased survival in older seedlings e.g. Melaleuca ericifolia
(Raulings et al., 2007), and woody species in Spain (Siles et al., 2010) which were
attributed to root depth and size at planting. These observations suggest there may be
a minimum size for seedlings to be planted, while Varone et al. (2012) demonstrated
that there is maximum seedling size for planting as well (i.e. three year old seedlings).
Therefore, pre-growing seedlings in nursery conditions for six months can allow roots
to grow long enough to explore soil layers after planting while minimising the chance
of producing root-bound seedlings.
Pre-translocation substrate only had a significant effect on survival of the older
seedlings planted in July. Opposite to our expectation, seedlings grown in potting mix
had greater survival and grew faster in the months after transplantation.
Subsequently, after the first summer (i.e. by March 2011) the survival difference
between the potting mix and sand grown seedlings remained constant. This suggests
that the potting mix eased ‘transplantation shock’ for young seedlings in the few
months after transplantation. A number of forestry studies have also found higher
survival of seedlings grown in potting mix compared with bare rooting, and this was
attributed to the undisturbed root system protected inside the potting mix plug. Roots
were able to maintain nutrient and water uptake when planted in this way, which led
to lower seedling water stress (Dixon et al., 1983; Grossnickle, 2005; Nilsson &
Örlander, 1995). The planting method for potting mix seedlings was also much easier
and faster. These seedlings could simply be ‘plugged’ into the transplantation holes,
whereas the seedlings grown in sand and planted with bare roots needed to be
carefully placed in holes which were refilled with native soil to ensure adequate root to
soil contact. The latter process also damaged part of the root system by breaking some
roots as the sand was being washed off (pers. obs.). In addition, for unknown reasons,
the sand-grown plants were significantly larger at transplant than the potting mix-
grown plants, for the older and larger batch A seedlings only. Thus, they would have
had a greater evaporative demand. As I suspect that their root systems would have
suffered greater damage during transplantation, this demand would have been
compromised by a reduced water transport capacity. This could explain why only the
Chapter 2
44
survival of batch A seedlings was significantly reduced when grown in sand. Therefore,
planting seedlings with a plug of potting mix is likely to lead to better establishment
and survival.
Surprisingly, watering treatments and microhabitat did not have a significant influence
on the survival of translocated seedlings. The majority of seedlings obtained sufficient
water to survive over the two year monitoring period, independent of watering and
microhabitat treatments. Seedlings experienced the greatest decline in survival one
month after watering began (December 2010 – Jan 2011), even though there was a
large rain period during this time. This rainfall was evident in the 20–60 cm soil depth
in January 2011 but had dissipated by March 2011 which suggests that this event was
short-lived in the soil profile. Despite substantially below-average rainfall during the
second summer, the vast majority of seedlings survived. By this time, seedlings may
have had longer root systems to obtain soil water and/or their cell functions were pre-
conditioned by the previous summer to survive the second, more severe summer
drought (Kozlowski & Pallardy, 2002).In addition, Acacia are generally assumed to be a
drought tolerant genus (Adams et al., 2010). This tolerance is partly ascribed to their
production of phyllodes instead of leaves. Phyllodes are altered petioles that have
been broadened and flattened which are thought to be an adaptation to drought due
to their greater water use efficiency and water storage capabilities (Brodribb & Hill,
1993; Pasquet-Kok et al., 2010; Warwick & Thukten, 2006). Furthermore, many woody
perennials in southwest Australian, including the Acacia species, have strong drought
tolerance. Some studies have shown that Acacia seedlings use osmotic adjustment,
while others demonstrate an adjustment of the flow rate through the xylem to avoid
drought-induced cavitation (Adams et al., 2010; Clemens & Jones, 1978; Otieno et al.,
2001). Therefore, the high survival of translocated seedlings is likely a reflection of
inherent drought tolerance.
Chapter 2
45
Water supplementation and seedling growth
In general, translocated seedlings that were watered on a weekly basis were
significantly taller compared with seedlings in plots with monthly and no water
supplementation. It was also observed that seedlings in weekly-watered plots
appeared to have larger more fleshy phyllodes especially when compared to seedlings
in plots without irrigation (pers. obs.).However, these differences in seedling
morphology did not correspond to significantly higher relative soil moisture in weekly
watered plots. The effect of watering was not evident in the upper soil layers (20–60
cm), and non-irrigated plots only had slightly lower relative moisture contents at
intermediate depth (60–100 cm). As soil moisture was measured a day before the next
watering cycle, these results suggest that water in more shallow layers was taken up
by the seedlings and the surrounding vegetation within a week. In contrast, the effect
of watering at intermediate depths lasted for at least a month. Thus, it is likely that
monthly watered seedlings would have experienced only a very short-term increase in
soil moisture in the upper layer (i.e. for less than a week), whereas the seedlings in
weekly watered plots would have received more regular increases in soil moisture.
Measures of transpiration also provided evidence that seedlings in plots with water
supplementation had a higher water supply and thus transpiration 24 hours after
watering relative to seedlings with no water supplementation. Higher transpiration
and a faster growth rate of seedlings in the weekly watered plots suggests that water
in the upper soil layers (<60 cm depth) was limiting growth over summer and that not
many roots penetrated to deeper depths during the two years of establishment. This
interpretation is consistent with what is known about soil beneath E. wandoo
woodlands which are often underlain by high bulk density clay layers that are difficult
to penetrate. Water supplementation during the first summer also had a strong effect
on the growth of roots and shoots of Pinus sylvestris seedlings in south- east Spain,
which allowed roots to explore a larger volume of soil during the second summer
without watering (Castro et al., 2005). Longer term monitoring will be required to
determine if weekly watered A. awestoniana seedlings maintain their growth
advantage, even after irrigation is turned off.
Chapter 2
46
The effect of microhabitat on seedling growth was generally dependent on seedling
batch. In most comparisons, seedlings placed in plots under E. wandoo were slightly
smaller. This could be due to competition for water and nutrients with the overstorey
trees, as E. wandoo is known to be relatively shallow-rooted and suppress
undergrowth (Lamont, 1985). Using sap flow recordings before and after a summer
rain event, Burgess (2006) found that while E. wandoo were principally using water in
deeper soil layers (up to 10 metres), this species was collecting water from upper
layers and re-distributing it to lower depths. Although there were no consistent
differences in soil moisture (20–100 cm depth) between microhabitat locations, effects
of irrigation were short-lived and measurements were taken after water had already
been taken up. Thus, possible differences between the habitat positions in the days
after watering would not have been detected. Apart from a competitive effect, the
reduced growth of seedlings placed under E. wandoo could also be related to the
lower light levels under the canopy and the significantly higher rates of herbivory.
Other studies have also identified greater herbivory in shaded areas (Collinge & Louda,
1988; Muth et al., 2008). This result could be due to shade plants using less energy on
herbivore defence and thus being more palatable and/or shade being a refuge for
herbivores, particularly during the hot, dry summer (Maiorana, 1981).
In contrast to the above observations, the weekly watered seedlings planted in August
(i.e. batch B) grew significantly faster in plots positioned under E. wandoo, suggesting a
possible facilitative effect. As these seedlings were 80% smaller than batch A seedlings,
it is likely they would have been much more vulnerable to the high temperatures and
drying of the soil surface in the more exposed open habitat positions. However, not all
batch B seedlings showed this response and especially seedlings in the un-watered
plots. These seedlings grew substantially taller in the open habitats. These contrasting
results may be due to a facilitation/competition trade-off (e.g. Sanchez-Gomez,
Valladares et al. 2006). For seedlings in the weekly watered plots, the facilitative
effects may have dominated as these seedlings were watered regularly, whereas the
effects of increased competition with E. wandoo roots may have prevailed in the non-
watered plots. Seedlings that were watered less frequently may have predominantly
allocated resources to root growth in response to their drier surface soils and have
Chapter 2
47
deeper roots (e.g. Poorter et al., 2012). Clearly, information on root distribution of
seedlings would be required to test these hypotheses further and better understand
the facilitative and competitive effects of surrounding vegetation.
Infrared thermography
The use of the infrared thermography was limited by the variability of measurements.
This variability is likely to be related to the frequent changes in the environmental
conditions during the measurements such as cloud cover and solar radiation, solar
angle, wind speed and humidity, rather than variability explained by the experimental
treatments. Ideally, measurements would be carried out over relatively short time
frame and under cloudless, windless conditions. Alternatively, climate variables can be
measured locally and can be used to correct for environmental changes. Past studies
using infrared thermography have incorporated detailed climatic measurements using
a climate station on site (Lenthe et al., 2007) and using wet and dry reference surfaces
to calculate leaf transpiration rates (Jones et al., 2002; Leinonen et al., 2006). Despite
the shortcomings, infra-red thermography has clear advantages as it is truly non-
invasive and can, unlike gas exchange measurements, be readily performed on small
seedlings. Although it has not been implemented in translocation studies thus far,
infrared thermography could complement traditional growth and survival
measurements by identifying those seedlings that require additional watering or
identifying optimal microhabitats for subsequent translocations.
Implications for translocations and threatened plant conservation
This study has identified a number of factors that can be readily manipulated to
improve the establishment of seedlings during translocations in a seasonally dry
environment. The following list includes considerations and recommendations for
future management that are based on this study.
Chapter 2
48
● Seedling size at transplant is one of the most important variables influencing
subsequent survival and growth. In seasonally dry environments, seedlings with a
larger root size are more likely to explore deeper soil layers for moisture faster thereby
avoiding desiccation.
● Planting seedlings in potting mix ‘plugs’ is much faster and reduces the chance of
root breakage during planting compared with planting seedlings with bare roots, which
subsequently improves initial survival and growth.
● Watering over summer may not always be required for higher initial survival,
particularly if there is a large rain event during the first summer since planting.
● In terms of seedling growth, it is beneficial to supply small amounts of water on a
weekly basis rather than a large amount infrequently in environments where water
availability in the upper soil layers is short-lived. This may be beneficial for long-term
survival as larger seedlings will be more competitive for soil water and nutrients.
● Planting seedlings close to canopy species may provide protection in some
circumstances but the canopy species may also be a source of competition and
herbivores can be more prevalent under canopies. Additional watering may be needed
to help seedlings overcome the competitive effects of canopy species.
● Infrared technology could be a useful tool for translocations, especially to identify
sources of stress (e.g. drought, shade, pathogens) of planted seedlings in seasonally
dry environments. This technology may ultimately help to improve seedling
establishment success and replaces more traditional and impractical tools such as the
porometer.
Chapter 3
49
Chapter 3
Seedling size and microhabitat influence survival
and growth of the Critically Endangered Banksia
ionthocarpa subsp. ionthocarpa in an experimental
translocation
Abstract
Banksia ionthocarpa subsp. ionthocarpa is one of eight Critically Endangered Banksia species
in southwest Australia. Past studies have shown that seedlings of this slow-growing species
are highly vulnerable to summer drought. This study aimed to advance our understanding of
safe site requirements for seedling establishment to enhance the long term translocation
success for this species. A key benefit of this study would be the increased size of a
previously translocated population and the improved likelihood of establishing a secure
population. I tested the effects of planting season (autumn versus late winter), pre-
translocation growth substrate (potting mix versus sand), watering frequency (weekly versus
monthly), and planting in different cleared microhabitats (bare, low open heath and taller
dense heath). The translocation was successful with 75% overall survival after two years.
Seedlings that were grown in potting mix and planted before the onset of winter rains were
up to 86% larger in size at planting and had significantly higher chances of survival. For
seedlings grown in sand prior to translocation, maximum root length was the most
important predictor of subsequent survival in the field. Microhabitat had a significant effect
on survival and growth, which was partly dependent on watering frequency and pre-
translocation growth substrate. Seedlings in tall heath habitats were consistently smaller
and more likely to have chlorotic leaves which were associated with thicker litter layer.
Seedlings that were watered monthly, grown in sand and planted with bare roots in this tall
heath habitat had the lowest survival. More frequent watering also improved growth and
survival in the most drought-prone and shallow-soil bare habitats. The most favourable
microhabitat for seedling survival and growth were plots in the low heath habitat which
were most similar to this species natural habitat. The low heath microhabitat also had the
greatest vegetation regrowth, suggesting effects of other plants were facilitative rather than
competitive. These results highlight the importance of manipulating microhabitats through
vegetation removal and different watering frequencies within translocations to identify safe
sites for seedling establishment.
Chapter 3
50
INTRODUCTION
It is estimated that 20% of global plant species are currently at risk of extinction (IUCN,
2013). In-situ conservation, through maintaining protected land and conservation
reserves, is an important prerequisite to preserve these species (Hoffmann et al.,
2008). Although in-situ measures are often effective, in some cases a combination of
both in-situ and ex-situ conservation approaches are required to prevent extinction
(Volis et al., 2011). Due to widespread fragmentation in the biodiversity hotspot of
southwest Australia, the conservation of the 153 Critically Endangered plant species in
this region is likely to require both in-situ and ex-situ approaches (Smith, 2013).
Translocations are one example of ex-situ approaches for management of threatened
species. This approach involves introducing species to a new suitable habitat or
augmenting existing populations using seedlings or seeds (Guerrant & Kaye, 2007).
Growing debate on the value and methodology of species “reintroductions” and
“translocations” suggests that this management approach is complex, as it requires an
understanding of the species ecological requirements (Guerrant, 2013; Sarrazin &
Barbault, 1996). For many threatened species this information is limited. However,
scientists and land managers continue to learn, through an adaptive management
process, from past translocation successes and failures. This process will ultimately
improve our understanding of threatened species ecology as well as developing
techniques that improve seedling establishment. For example, many recent
translocations have used seedlings that have been germinated ex-situ as an increasing
body of literature suggests higher establishment rates with planted seedlings
compared with direct seeding (Albrecht & McCue, 2010; Jusaitis et al., 2004). This
example highlights the importance of reporting translocation failures as well as
successes, which both help to develop our understanding of the ecological
requirements for successfully establishing translocated populations of threatened
species (Kennedy et al., 2012; Monks, 2008).
Manipulating microhabitats has become a tool to identify potential safe sites for
translocated seedlings (Godefroid et al., 2011; Guerrant, 2012). The safe site
hypothesis predicts that seeds will land in a heterogeneous landscape and
Chapter 3
51
germination, establishment and survival to maturity will depend on the relative
favourability of the abiotic and biotic conditions (Harper, 1977). Creating safe sites by
manipulating microhabitats within translocations has improved understanding of the
factors which effect seedling survival and growth (Jusaitis, 2005; Roncal et al., 2012).
Some experimental translocations have created different microhabitats by removing
competition with surrounding vegetation (Morgan, 1999; Pavlik et al., 1993), planting
in vegetation/soil types that are different to those of natural populations (Lawrence &
Kaye, 2009; Montalvo & Ellstrand, 2000; Raabová et al., 2007) and using disturbances
such as grazing and litter removal (Burmeier & Jensen, 2009; Gordon, 1996). Using a
range of microhabitats also maximises the chances of transplanting seedlings into at
least one favourable environment as well as gaining knowledge about a species’
ecology and its establishment requirements.
Although manipulations at the microhabitat level have been commonly incorporated
into the practice of translocation, pre-translocation treatments at the seedling level
have not been often used (Guerrant, 2012). Different protocols used in nurseries
during germination and early growth could potentially affect the viability of these
nursery-grown seedlings when planted in the field. For example, growing seedlings in
substrates with relatively high nutrient contents and ample water, although optimal
for growth, is likely to generate seedlings with artificially high shoot to root ratios
(Poorter et al., 2012). In Mediterranean climates, such as southwest Australia, this
root: shoot ratio could lead to excessive evaporative demand of planted seedlings and
a higher vulnerability to desiccation. Roots of these seedlings may also avoid, at least
initially, the drier native soil which could potentially delay establishment. This delay is
significant because seedlings have a short window to establish before the onset of the
long summer drought. One way to overcome this potential problem is to plant
seedlings directly into field soil. This would force bare roots into immediate contact
with field soils as a growth medium. Apart from growth substrate, timing of planting
can also influence survival. Planting before or at the onset of the winter rainfall,
combined with efforts to irrigate the seedlings over their first summer drought, would
potentially allow roots, and thus seedlings, to become well established (Albrecht &
McCue, 2010).
Chapter 3
52
Southwest Australia is a biodiversity hotspot with a very high proportion of endemic
plant species vulnerable to a range of threatening processes such as major habitat
fragmentation and altered disturbance regimes (e.g. infrequent fire), weeds, grazing by
introduced animals, Phytophthora dieback and altered hydrology (Hopper & Gioia,
2004; Myers et al., 2000). The region contains many naturally rare species that are
often restricted to specific soil types (edaphic specialists). Many species also have
short dispersal distances and long life cycles which makes them vulnerable to rapid
environmental change (Cowling et al., 1996; Hopper, 1979; Hopper, 2009). As a
consequence of widespread disturbance and habitat clearing, these naturally small
populations become increasingly isolated with inbreeding potentially compromising
future reproduction (Llorens et al., 2012). With these constraints in mind, ex-situ
conservation through translocations has become increasingly important to reduce the
extinction risk for flora in southwest Australia. Since 1998, approximately 105
translocations have been established for various genera throughout the region with
varied rates of success (Coates, D. pers. comm., January 2011).
Several on-site treatments have been applied throughout southwest Australia to
improve translocation success such as mulching, shading, mounding and ripping
(Cochrane et al., 2000; Monks, 2008). One microhabitat manipulation that has not
been explored within this region is removing or reducing the aboveground biomass at
the planting site prior to planting seedlings for translocation. Elsewhere, results from
past translocations have demonstrated that this method generally increases water
availability due to a reduction in competition with surrounding vegetation (Buisson et
al., 2008; Gordon, 1996). In fire-prone regions such as southwest Australia, many
species regenerate after a fire in bare patches where mature vegetation has been
scorched (Bell, 2001). Thus, seedlings often establish in an environment with low
competition from mature vegetation for both water and nutrients. While it is not
always feasible to use fire to reduce aboveground competition in a conservation
setting, it is possible to create gaps by manual removal of vegetation (Petru & Menges,
2003).
Chapter 3
53
Over 90% of all Banksia species occur in southwest Australia including many that are
rare and geographically restricted (Witkowski & Lamont, 2006). Banksia species are
usually found on nutrient-poor soils and tend to have limited dispersal (Merwin et al.,
2012). Approximately 60% of Banksia species in the region are reseeders and
bradysporous (i.e. retain their seeds for a number of years in woody fruits within their
canopy), and it has been proposed that fire may be a key factor for local dispersal and
recruitment (Lamont et al., 2007). Currently, there are eight Critically Endangered
Banksia species and translocations have been attempted for four of these species: B.
anatona B. montana, B. brownii and B. ionthocarpa subsp. ionthocarpa between 2000
and 2010 (Smith, 2013). Water supplementation over summer was applied to all
translocations, which presumably partly explains the high survival of B. anatona, B.
brownii and B. montana seedlings (Barrett et al., 2011; Center for Plant Conservation,
2013). However, survival of translocated Banksia ionthocarpa was low (~20%) after
two years (Monks, 2008).
Here, I aimed to further understand the microhabitat requirements and seedling
characteristics for Banksia ionthocarpa subsp. ionthocarpa by establishing an
experimental translocation. I incorporated experimental treatments that had been
used previously (i.e. water supplementation) and additional treatments that had not
been implemented in any translocations within southwest Australia. These novel
treatments included planting seedlings in three microhabitats (bare soil, low and tall
heath) and at two planting dates (autumn or late winter). In each of the microhabitats,
aboveground vegetation was removed prior to planting to simulate the effects of fire,
and to reduce competition for water and nutrients. It was hypothesised that seedlings
planted into tall heath would have higher survival and growth as the vegetation (prior
to clearing) was dense suggesting higher water availability. It was expected that weekly
instead of monthly watering would further improve seedling growth and survival. In
addition, seedlings were grown in two substrates— sand or potting mix—prior to being
planted in the field. The latter substrate was used to test the effect of planting
seedlings with their nursery growth medium (conventional: potting mix), and without
it by removing the sand from the root systems prior to planting. I hypothesised that
roots planted straight into the field soil would grow faster, thereby facilitating seedling
Chapter 3
54
establishment. The survival and growth of the translocated seedlings were monitored
for two years after planting and these data were correlated with environmental
conditions (i.e. local rainfall, soil moisture and temperature) to help explain the results.
METHODS
Study species
Banksia ionthocarpa subsp. ionthocarpa (hereafter B. ionthocarpa) is ranked as a
Critically Endangered species in Western Australia under Criteria B1ab(v) and B2ab(v)
of the Red List prepared by the World Conservation Union (IUCN, 2012). This prostrate
Banksia species grows in dense tufts with short stems and leaves up to thirty
centimetres long. Banksia ionthocarpa persists in one remnant population (two sub-
populations) that is confined to open low heathland on shallow soils over spongelite
and surrounded by agricultural land (Figure 1). Sub-population 1A has 450 mature
individuals and is found on a recreation reserve while sub-population 1B has 203
mature plants and is located on public utility land (population numbers as of 2012 –
Barrett, S pers. comm. July 2013). The total area of occupancy is approximately five
square kilometres. Both sub-populations are found in a long unburnt habitat
(approximately 25 years) and the numbers of plants have declined by 30% in seven
years (Department of Conservation and Land Management, 2004). This decline is
thought to be due to senescence, but weed invasion, drought and susceptibility to the
soil pathogen Phytophthora cinnamomi may potentially impact the health and survival
of the remaining individuals (Department of Conservation and Land Management,
2005).
Banksia ionthocarpa is a slow-growing species that produces little seed and whose
adults are killed by fire (Millar et al., 2010). Seedlings that recruit after fire can have
high mortality (i.e., 70% of recruits died), probably due to the drying soil conditions
over summer (Department of Conservation and Land Management, 2004; Monks,
1999). Both sub-populations are found on public land and thus generally have less
protection than in conservation reserves. Therefore, a translocated population was
Chapter 3
55
established in 1999/ 2000 in a nearby Nature Reserve with a similar habitat to the
source population. Translocated seedlings were shaded to reduce water loss and
fenced to exclude herbivores. Although these seedlings were watered for the first two
summers after planting, by 2004 only 20% were recorded as having established at the
site (Department of Environment and Conservation, 2008). The results of this
translocation suggested that watering frequency and microhabitat type might be
important factors contributing to the survival and growth of translocated seedlings.
Figure 1 – Location of the translocated B. ionthocarpa population in respect to the nearest
climate station and natural sub populations 1A and B. Coordinates are not provided due to the
requirement not to disclose the exact locations of the Endangered species.
Germination
The Department of Parks and Wildlife, Western Australia has collected seeds of B.
ionthocarpa since 1999, which have subsequently been stored in -20 °C at the
Threatened Flora Seed Centre in Perth, Western Australia (Crawford, A pers. comm,
June 2011). Seeds from both sub-populations were used to grow seedlings for this
study and were sourced from 46 and 47 maternal plants from sub- populations 1A and
1B respectively (differences in survival and growth were not clearly differentiated by
maternal population and thus not reported in this study). Seeds were germinated at
the Threatened Flora Seed Centre, on two dates (batch A = 12th January 2010, batch B
Chapter 3
56
= 8th April 2010) to ensure seedlings of similar size were available to plant in the early
and late planting treatments. Seeds were soaked for 15 minutes in a 50% solution of
Plant Preservative Mixture (Plant Cell Technology, Washington, D.C.) to surface-
sterilise the seeds before being placed onto agar in Petri dishes (7.5 g agar per L
distilled water). Petri dishes were placed in a constant temperature room set to 15 °C
with a 12 hr light/12 hr dark photoperiod to stimulate germination. After germination,
seeds were transferred to small cell trays. These trays were transported to Kings Park
and Botanic Garden Nursery, Perth, where they were planted into 50 × 50 × 125 mm
pots. Seedlings from batch A were planted into two different soil substrates (potting
mix or sand). The potting mix was the standard Kings Park mix which consisted of
composted jarrah sawdust, river sand and coarse sand in a ratio of 2:1:0.5. Osmocote
slow release native fertiliser (low phosphorus plus trace elements – 3 kg per cubic
metre), ferrous sulphate (500 g/m3), dolomite (2.3 kg/m3) and lime (2.5 kg/m3) were
added to the potting mix. The sand substrate consisted of washed river sand and
Osmocote slow release native fertiliser at 3 kg/cubic metre. Seedlings from batch B
were all planted in potting mix. Seedlings were allowed to mature and acclimatise
outdoors at the Kings Park Nursery for four months over autumn/ winter 2010.
Experimental Design
The translocation site was located in the Kalgan Plains Reserve, Western Australia,
approximately five kilometres northwest of the remnant population (Figure 1). The site
was chosen because the soil type (shallow gravel-clay over spongolite) and open
heathland vegetation with associated species such as Taxandra spathulata, Melaleuca
pentagona, Hakea marginata, Neurachne alopecuroidea and Opercularia spermacocea
was similar to the natural habitat (Hocking, 2011). Eighteen 3.5 m × 3.5 m fenced plots
were randomly allocated to one of three microhabitat types (bare soil, low open heath
and tall dense heath; n = 6 plots per habitat type). The vegetation in the low and bare
heath plots was most similar to the natural habitat of B. ionthocarpa (Hocking, 2011).
The plots occupied an area approximately 75 m × 50 m. Fences were built to protect
seedlings from grazing by mammal herbivores. Vegetation in the low and tall heath
Chapter 3
57
plots was removed to ground level. The fresh biomass of the vegetation removed from
each plot was recorded on site. Batch A seedlings were planted on the 29th and 30th of
April 2010. These seedlings comprised two thirds of the total number of translocated
seedlings and included seedlings grown in sand and potting mix. The final third of
seedlings (batch B; all potting mix) were planted on the 19th and 20th of August 2010.
Seedlings were planted at 30 cm intervals in a rectangular grid measuring 120 x 150 cm
within each plot. Seedlings from both batches and soil substrates were randomly
allocated to positions on the grid. Seedlings grown in potting mix were planted
together with the plug of soil attached to their roots. For seedlings grown in sand, the
sand was washed off the roots before planting and root length (mm) was measured
before planting to quantify the relationship between root size and seedling survival.
Seedlings were planted within small depressions created to increase local water
availability (Whisenant et al., 1995).
Two watering treatments were used to test the effect of frequency of summer
watering on seedling growth and survival (Table 1). Drip irrigation was applied to half
of the plots in each of the microhabitat types in November 2010, with a dripper for
each individual plant connected to a 4,500 L water tank and solar pump. The tank was
filled with chlorinated drinking water on a monthly basis over summer. Each plant in
the drip irrigated plots received approximately one litre of water each week just prior
to dawn as per the current standard for plant translocations in Southwest Australia
(Dillon, R pers. com, November 2010). The plots designated to be hand-watered were
watered on a monthly basis with 120 L of water from 7 am – 9 am in the morning. This
volume of water was based on one litre for the thirty plants in each plot over four
weeks. Both watering treatments ran over two summers from December 2010 to
March 2011 and from January 2012 to March 2012, water was not applied in
December 2012 due to the occurrence of a large rainfall event. A non-watering
treatment was not included in the design given that previous studies demonstrated
that seedlings of this species tend to be susceptible to drought. There was also a need,
from a conservation perspective, to maximise seedling establishment given the limited
seeds available for this species. The current recommendation is to water translocated
Chapter 3
58
seedlings for up to two years after planting, which is applied in the case of each new
translocation in a precautionary manner (Monks, L pers. com, November 2013).
Table 1 – Summary of experimental treatments at the plot and seedling level within in the
translocation of B. ionthocarpa
Experimental Treatment Treatment Levels
Plot treatments 1. Bar
Water supplementation 1. Weekly watering (drip irrigation) 2. Monthly watering (by hand)
Microhabitat type 1. Bare soil 2. Low heath 3. Tall heath
Seedling treatments
Soil substrate 1. Potting mix 2. Sand
Planting season 1. Early (April 2010) 2. Late (August 2010)
Monitoring of Seedlings
The translocation site was visited every month in summer (i.e. January to March) and
once per winter between April 2010 and July 2012. Seedling survival was recorded and
the above ground volume of seedlings was approximated using the formula for an
ellipsoid: v= (∏/3)r2h (r=longest width and h=height) (Ludwig et al., 1975). Volume was
measured six times during the monitoring period. Monks (1999) observed that mature
B. ionthocarpa changed foliage colour over seasons (Figure 2) and that chlorotic plants
generally had significantly lower xylem water potentials which is indicative of drought
stress (Nardini et al., 2013). Therefore, leaf chlorosis was quantified by recording the
number of plants with yellow leaves per plot.
Chapter 3
59
Figure 2 – Banksia ionthocarpa seedlings, left – healthy green leaves and right – chlorotic
seedling.
Soil moisture was measured to six centimetres depth using a soil moisture probe
(MPM-160-B 12 bit resolution, ICT International, NSW, Australia) to determine if
watering frequency altered available water in the upper soil layers. Soil moisture was
measured at five random locations within each plot one week after drip irrigation and
one month after hand-watering throughout the monitoring period. The relative
difference in soil moisture (millivolts) was calculated before (i.e. October 2010) and
after watering to compare treatment effects at the plot level.
Percentage cover of regrowth and weeds was estimated over the first summer after
planting (Jan – Mar 2011) to quantify the potential severity of competition within
plots. In January 2011 and 2012, the number of B. ionthocarpa seedlings with regrowth
(seedlings, grasses or resprouting vegetation) in their immediate vicinity (within the 15
cm diameter planting depression) was also quantified.
Small round temperature loggers (17 mm diameter; 1-wire Thermocron I-buttons, San
Jose, USA) were used in January 2012 to identify differences in soil surface
temperature according to the treatment combinations (i.e. microhabitat × watering
treatments). Thirty temperature loggers were placed at one centimetre depth next to
randomly chosen seedlings (two plots per microhabitat / watering combination). These
Chapter 3
60
loggers were programmed to collect temperature at three hour intervals per day for
one month.
Statistical analysis
To analyse the effect of plot variables on survival and growth (volume) of Banksia
ionthocarpa seedlings during the monitoring period, linear mixed models were
constructed in the nlme package in R (Table 2; R Core Team, 2013). Seedlings from the
three substrates and planting season combinations (i.e. batch A/ potting, batch A/
sand and batch B/ potting) were analysed in separate models due to the complexity of
the experimental design and large variation between for survival and growth. Seedling
volume was also log10 transformed to account for variation within each of these
combinations (Zuur et al., 2007). Maximal models were created within a Maximum
Likelihood (ML) framework with survival or log10 growth as response variables and
plot treatments (watering and microhabitat) as explanatory variables (including all
interactions). Plot was included as a random factor in each model. The maximal model
was subsequently reduced using stepwise deletion and a chi-squared ANOVA was
employed to compare Akaike Information Criteria (AIC) of maximal vs. reduced
models to find the best correlation between variables (Bolker et al., 2009). Diagnostic
plots were used to detect heterogeneous variances by plotting residuals against each
explanatory variable. If these were detected, regression slopes were then allowed to
vary with the heterogeneous variable. The optimal models were then analysed using
Restricted Maximum Likelihood (REML) which incorporates the random effect within
model predictions for a more accurate analysis (Bolker et al., 2009).
The size of seedlings at planting and their subsequent survival after two years were
compared among planting seasons and substrates using a t-test with unequal
variances. Seedling survival at July 2012 was regressed with presence of regrowth in
immediate vicinity of the seedling in January 2011 and 2012. Survival of seedlings
grown in sand was also correlated to root length at planting (Table 2).
Chapter 3
61
Table 2 – Response variables used to characterise the effects of different treatments at
seedling and plot level within the translocation of B. ionthocarpa seedlings. These variables
were measured at different times during the two years of monitoring. Separate models were
also created based on the three combinations of pre-translocation substrate and planting
season for seedling survival and growth. The responses were then included in separate linear
mixed models with all interactions between treatments initially included and plot as a random
variable. Slopes of the model were allowed to vary when heterogeneous variances were
detected.
Response variable Period of measurement Treatment variables
included in model
Model
distribution
Seedling survival November 2010 to July 2012
1. Plot watering frequency
2. Plot microhabitat 3. Centred initial
seedling volume
Binomial
(survival = 0,
dead = 1)
Log10 seedling size Initial size at planting (April or August 2010)to July 2012
1. Plot watering frequency 2. Plot
microhabitat
Normal
Leaf chlorosis June 2010 June 2011 March 2012 July 2012
1. Planting season 2. Pre planting soil
substrate 3. Plot watering
frequency 4. Plot microhabitat
Binomial
(green = 0,
chlorotic = 1)
Soil moisture From December 2010 to February 2012
1. Plot watering frequency 2. Plot microhabitat
Normal
Chapter 3
62
RESULTS
Climate over monitoring period
The translocation site experienced below-average rainfall for most of the 26 month
monitoring period (Figure 3). January 2011 received nearly 75% more rainfall than
average and there was also above-average rainfall in spring to early summer 2011
(October to December).
Figure 3 – Observed monthly total rainfall (mm) over the experimental period,
between April 2010 and July 2012 from White Gums climate station (1km from
translocation site) compared to long term total rainfall (1920 – 2009). Arrows indicate
the two translocation planting dates in 2010.
Summer rainfall during past (2000) and present translocation
During the first summer after planting, both translocation attempts received above
average rainfall in January (61.4 cm in Jan 2000 vs. 67. 4 cm in Jan 2011). However, the
second summer of 2000 – 2001 was very dry (total rainfall 27.4 cm Dec – Feb)
0
20
40
60
80
100
120
Apr-10 Jul-10 Oct-10 Jan-11 Apr-11 Jul-11 Oct-11 Jan-12 Apr-12 Jul-12
Rai
nfal
l (m
m)
Month
Observed monthly total
Long term monthly total
Late planting (batch B)
Early planting (batch A) First summer Second summer
Chapter 3
63
compared to the rainfall conditions in the current translocation (total rainfall 83.4 cm;
Dec 2011 – Feb 2012).
Characterisation of microhabitat types
a) Competition from surrounding vegetation
The amount of fresh biomass that was cleared from each plot before planting the
seedlings differed greatly among microhabitat types. Tall plots had the most biomass
which likely reflected the higher soil water availability in these microhabitats (Table 3).
Table 3 – Average fresh aboveground biomass that was removed per plot ± SE (n = 6) for each
microhabitat type prior to planting B. ionthocarpa seedlings.
Microhabitat type Fresh biomass (kg) Vegetation type
Bare 2 ± 0.4 Protruding rocks and bare soil with sparse grasses
Low 13 ± 2.3 Open shrubland with sparse grasses up to 1 metre high
Tall 97 ± 3 Dense shrubland up to 1.5 metres high
Regrowth in plots, i.e. new seedlings, resprouting shrubs and grasses, was first
observed in November 2010 (seven months after the vegetation had been cleared) and
was quantified in the first summer after planting (January 2011). Bare heath plots
often had a moss layer on the ground, particularly on rocky surfaces. Some low heath
plots contained up to 70% grass and sedge cover and both low and tall heath plots had
10–20% woody cover (mostly myrtaceous species). During the second summer, nearly
a quarter of B. ionthocarpa seedlings had regrowth in their immediate vicinity (15 cm
diameter around the plant) and thus exposed to potential competition from regrowth.
A slightly higher proportion of seedlings in low heath plots (28%) had this regrowth
compared to bare and tall heath plot (18% and 21% respectively). In the low heath
plots, it was found that the presence of regrowth in summer significantly increased the
chance of survival at the end of the monitoring period (regrowth present first summer
p = 0.04; second summer p = 0.03).
Chapter 3
64
b) Soil moisture
Relative changes in soil moisture content of the top 6 cm reflected seasonal rainfall
trends for all plots with a decline towards the end of both summers and an increase
during winter (Jun 2011; Figure 8). Before water supplementation commenced in
December 2010, most plots had higher soil moisture contents compared to their
October 2011 values, with some differentiation amongst plots that was not clearly
related to habitat type (Figure 4). Subsequently, the relative soil moisture content was
consistently higher in bare plots compared to low and tall heath plots (March 2011
d.f.15. bare vs. low heath p = 0.02, bare vs. tall heath p = 0.004; February 2012 d.f.15.
bare vs. low heath p = 0.03, bare vs. tall heath p < 0.001). There was no effect of
watering frequency on relative soil moisture contents measured one week or one
month after the last watering.
Figure 4 – Percentage change in soil moisture ± SE in the top 0–6 cm relative to pre-watering
(October 2010) measurements as dependent on plot watering frequency (monthly vs. weekly
over summer) and microhabitat type (Bare = bare soil, Low = low open shrubland and Tall =
dense shrubland) at the B. ionthocarpa translocation site. Average pre-watering moisture (mV
signal of probe) based on plot type in October were as follows (162 ± 9 weekly/ bare, 90 ± 9
weekly/ low, 167 ± 15 weekly/ tall, 129 ± 8 monthly/ bare, 99 ± 12 monthly/ low and 153 ± 18
monthly/ tall)
-80
-60
-40
-20
0
20
40
60
Dec-10 Mar-11 Jun-11 Sep-11 Dec-11
Perc
enta
ge c
hang
e in
soi
l m
oist
ure
Monitoring month
Weekly-Bare
Weekly-Low
Weekly-Tall
Monthly-Bare
Monthly-Low
Monthly-Tall
Chapter 3
65
c) Soil temperature
Soil temperatures were significantly lower in plots that had been watered twelve hours
prior (i.e. plots assigned to weekly watering at 5 am) throughout summer. Week one
was the anomaly in this trend as monthly watering was carried out 24 hours before
measurements this week (Table 4). These differences in soil temperature were not
observed 48 hours after watering (data not shown). There were no significant
differences in soil temperatures among habitats. However, week three (late January)
was the warmest in the month and there was a trend for soil temperatures to be
higher in bare and tall heath plots (~50 °C and ~53 °C respectively) that had not been
watered compared to the low heath plots (~45 °C).
Table 4 – Weekly time course of average maximum (2pm) soil temperature ± SE (°C) at one
centimetre depth next to B. ionthocarpa seedlings. Values are provided for weekly watered
plots, that had been watered 12 hours before each week’s measurement (n = 14), and for
monthly watered plots which were only watered 24 hours before the first week’s
measurement (n = 16). Soil temperatures among microhabitats were similar and therefore
combined over a four week period during the second summer since planting (January –
February 2012). Soil temperature was collected on a daily basis for one month. Statistical
comparisons were made within each week and stars denote significantly higher soil
temperatures (p < 0.01).
Week 1 Week 2 Week 3 Week 4
Weekly watered 40 ± 0.7* 30 ± 0.4 40 ± 1 32 ± 0.7
Monthly watered 36 ± 0.5 33 ± 0.3* 51 ± 1* 39 ± 0.6*
Seedling survival
There was a high overall survival (75%) of B. ionthocarpa seedlings at the end of the
two year monitoring period (July 2012). A month before the onset of water
supplementation, seedling survival was ~90% with no differences amongst treatments.
For both batches, the highest proportion of seedlings died during first six to seven
months since planting (batch A: 9% of total mortality occurred April – November 2010;
batch B: 17% of total mortality occurred August 2010 – March 2011).
Chapter 3
66
Planting season and pre-translocation substrate affected size at planting which
significantly influenced overall survival (Appendix 4). Seedlings grown in potting mix
and planted early in the season (batch A) were largest at planting and had the highest
survival after two years (Table 5). In contrast, seedlings planted at the same time but
grown in sand were ~30% smaller at planting and had significantly lower overall
survival rates. The seedlings planted later in the season (batch B), that were grown in
potting mix, were ~90% smaller than batch A seedlings. Despite their much smaller size
the batch B seedlings still had overall survival rates that were similar to the seedlings
grown in sand. Generally, the largest seedlings in each planting season and pre-
translocation substrate combination had a higher chance of survival (Table 5).
Table 5 – Survival (%) and initial size (mm3) of translocated seedlings of B. ionthocarpa as
dependent on pre-translocation substrate (potting mix or sand) and planting season (batch A:
April 2010, batch B: August 2010). Sizes are differentiated for seedlings that survived the
experimental period and for those that died. Values indicate means ± SE. Lower case and
capital letters denote significant differences in seedling size and percentage survival,
respectively. Sample sizes are provided in brackets.
Substrate Season Size alive
(mm3) Size dead
(mm3) Survival (%)
Potting April 2010 6902 ± 565a 4798 ± 1179
ac 84 ± 7
A (159)
Sand April 2010 5521 ± 646ab
2717 ± 474c 67 ± 10
B (105)
Potting August 2010 954 ± 73d 623 ± 77
e 68 ± 8
B (87)
Washing the roots of the seedlings grown in sand immediately prior to planting
enabled me to measure the maximum root length of each seedling. Seedling survival
was significantly associated with maximum root length (Figure 5). Seedlings with root
lengths less than five centimetres at planting had a significantly lower chance of
survival compared to seedlings with root lengths between 10–20 cm. Seedlings with
roots over 10 cm generally had a high chance of survival with 65–80 % survival after
two years. Moreover, seedlings with root lengths over 15 cm showed consistent high
survival during the experimental period, whereas seedlings with root lengths 10 – 15
cm showed a gradual decline. Although, as expected, many of the seedlings with long
roots were simply the largest individuals (based on shoot volume; Figure 6), a
Chapter 3
67
significant number were not among the largest individuals and still had greatly
increased survival chances compared to larger seedlings with shorter roots (Figure 5).
Figure 5 – Percentage survival ± 1 S.E. of seedlings grown in sand substrate as dependent on maximum root length at planting. Roots were categorised in four length classes: less than 5 centimetres (n = 7), 5–10 cm (n = 22), 10–15 cm (n = 93), 15–20 cm (n = 27). Different letters indicate 5% significance (separate analysis for each month).
Figure 6 – Initial size of B. ionthocarpa seedlings grown in sand before translocation correlated with their root length at planting in April 2010.
0 10 20 30 40 50 60 70 80 90
100
Nov-10 Mar-11 Jan-12 Jul-12
Perc
enta
ge s
uriv
al
Monitoring month
<5 cm
5 - 10 cm
10 - 15 cm
15 - 20 cm
R² = 0.16
-5000
0
5000
10000
15000
20000
25000
30000
35000
0 5 10 15 20 25
See
dlin
g vo
lum
e (
mm
3)
Root length (mm)
a a b b a b c c a b c c a a b b
Chapter 3
68
The survival of batch A seedlings was not significantly affected by the treatments at
plot level, although seedlings in weekly watered plots generally had higher survival
than monthly watered seedlings. Survival of seedlings grown in sand was more
variable, with generally higher survival for plants in weekly watered plots, and the
lowest survival in tall heath plots that were watered on a monthly basis (~40%; Figure
7).
Despite having higher mortality, the seedlings planted in late winter (batch B; all
potting mix) had a similar relative response to the plot level treatments as the
seedlings planted earlier with potting mix. Over the first summer, there was 38%
decline in survival for seedlings planted in monthly watered bare plots, which was
significantly lower than seedlings planted in the tall heath plots (15% decline, Appendix
4; Figure 7). Also, there was a trend that watering had a positive effect on survival in
both the bare and low heath habitats but not in the tall heath habitats.
69
Chapter 3
Figure 7 – Plot survival (%) ± SE (n= 3 plots per treatment combination) of translocated B. ionthocarpa seedlings as dependent on pre-translocation
substrate and planting season (batch A – April 2010 and batch B – August 2010) over the two year monitoring period. Seedlings were watered
between January and March during both years. Different symbols and lines indicate the effects of summer watering frequency (weekly or monthly)
and habitat type (Bare = bare soil, Low = low open shrubland and Tall = dense shrubland)
Chapter 3
70
Chapter 3
Seedling growth
Seedling growth was significantly influenced by microhabitat type which was first
observed a year after planting. Seedlings in tall heath plots had a significantly smaller
volume compared with seedlings in low and bare heath plots regardless of planting
season or pre-translocation substrate (Appendix 5). Watering did not have a consistent
influence on seedling growth throughout the monitoring period. In fact, some of the
largest seedlings measured in the second summer were located in monthly watered,
low heath plots (Figure 8). Regardless of planting season or substrate, seedlings in bare
plots that were monthly watered tended to have lower survival rates, and also had
reduced growth during at least a part of the experimental period (Figure 8).
Seedling chlorosis
Within a month after planting, 7% of seedlings in batch A displayed chlorotic leaves
and 83% of these seedlings were in tall heath plots (April to June 2010; other 7% were
located in low heath plots). At the end of summer, a high proportion of seedlings were
chlorotic and this was significantly more common in monthly-watered plots (45% of
seedlings in weekly-watered plots and 65% in monthly-watered plots p = 0.02).
Chlorotic leaves was less frequent overall in the two winters during monitoring but
significantly more common in tall heath plots compared to low and bare plots (June
2011 = 3, 4 and 30%; p<0.001 and July 2012= 6, 5 and 43%; p<0.0001 of seedlings in
bare, low heath and tall heath plots, respectively). The seedling level treatments
(planting season and pre planting substrate) did not have any effect on the occurrence
of chlorosis.
71
Chapter 3
Figure 8 – Growth of translocated B. ionthocarpa seedlings over time as expressed by changes in seedling volume ± SE (log10), and as dependent on pre-translocation substrate (potting or sand) and planting season (April 2010- batch A or August 2010- batch B). Different symbols and lines indicate the effects of watering frequency over summer (weekly or monthly) and habitat type (Bare = bare soil, Low = low open shrubland and Tall = dense shrubland) (n = 3 plots per treatment combination).
Chapter 3
72
DISCUSSION
There was substantially higher overall survival (75%) of the translocated B. ionthocarpa
seedlings after two years compared with the 23% survival rate after two years of
seedlings planted in the initial translocation at the same site (Department of
Conservation and Land Management, 2004). This difference could partly be attributed
to differences in summer rainfall. But as the highest proportion of seedlings death
occurred in the first six months to year of planting, other factors probably contributed
to higher survival in the current study. Incorporating an experimental approach with
seedling and microhabitat treatments was likely an important aspect. This approach
was also essential in disentangling the main factors influencing seedling survival and
characteristics of a safe site for this species. Shoot and particularly root size at
planting appear to have played a dominant role in the seedling survival of B.
ionthocarpa with larger seedlings more likely to survive after the first year. Significant
growth differences were observed among microhabitats, with seedlings in tall heath
plots remaining smaller and often displaying chlorosis. The low heath plots and the
more frequently watered bare plots were the most favourable habitats in terms of
survival and growth and these microhabitats were most similar to the remaining
natural habitats of B. ionthocarpa. These findings suggest that microhabitat differences
are a significant contributing factor to seedling survival and can potentially be used to
guide future translocations of this species and could be considered more broadly in
threatened plant translocation programs.
Planting season and seedling size
Differences in seedling size and planting season influenced the survival of seedlings
grown in potting mix. The seedlings planted in late autumn (batch A) were larger
(~85%) and had significantly higher survival compared to those planted in mid-winter
(batch B). The initial size difference between batches was most likely due to substantial
differences in nursery conditions, i.e. temperature and radiation in the pre-
translocation period, as batch A plants were grown over summer and batch B plants
Chapter 3
73
over late autumn to winter. I was not able to disentangle the effects of size and
planting season as there was no overlap seedling sizes between both batches.
However, within each of the three combinations of planting season and substrate, the
seedlings that survived after two years were generally taller than those that died.
Therefore, seedling size is likely to be important for survival as was observed in other
translocations such as Acacia awestoniana (Chapter 2) and Amorpha herbacea, an
endangered shrub that is restricted to limestone outcrops in South Florida
(Wendelberger et al., 2008).
The results for the seedlings grown in sand strongly suggest that maximum root length,
rather than overall plant size, is the major determinant of subsequent summer
survival. This is not surprising as sufficient root growth to access water at depth and
establishing good root/soil contact in the early stages of seedling establishment is
integral to avoid desiccation over summer (Grossnickle, 2005; Padilla et al., 2007;
Padilla & Pugnaire, 2007; Reader et al., 1993; Williams et al., 1997). This study suggests
that a root length of over 15 cm at planting will lead to high (~90%) and consistent
seedling survival over two years since planting. This can be achieved through either
growing bigger plants and/or planting early in winter or late autumn to allow for a
larger period of sustained root growth.
Soil substrate
Apart from seedling size, pre-translocation growth substrate significantly influenced
seedling survival, with seedlings planted with a plug of potting mix having improved
chances of survival compared to those planted with bare roots. A similar result was
obtained in Chapter 2 for Acacia awestoniana. These data suggest that root damage
that was sustained by washing sand off the roots and re-filling the planting holes
cancelled out any potential advantages of planting the roots directly into field soil.
Chapter 3
74
Interactions at the plot level
The placement of seedlings in different microhabitats significantly affected their
chances of survival and their growth rates. However, results were partly dependent on
substrate and watering frequency.
A) Bare plots
In the very shallow soils (~12 cm; Hocking, 2011) of bare habitats, the seedlings in the
weekly watered plots generally had higher survival and growth rates compared with
monthly watered seedlings, independent of growth substrate or seedling batch. This
result is not surprising as shallow soils in seasonally dry environments are amongst the
most drought-prone habitats (Poot et al., 2012). Indeed, an attempt by Monks (1999)
to translocate B. ionthocarpa seedlings into bare soil found only a 0–3% survival rate
after nine months which was hypothesised to be due to high soil compaction after
clearing. However, the relative soil moisture (0–6 cm) measurements were high for
bare plots regardless of watering frequency. This suggests that higher moisture was
retained in the topsoil of bare habitats compared with low or tall heath due to a lack of
plant biomass. But within 48 hours of watering, initial differences in soil temperatures
between watering treatments disappeared which suggested that the the cooling effect
of water in the upper six cm had largely dissipated. It is likely that the majority of
supplemented water would have infiltrated into deeper soil layers or would have been
lost through evapo-transpiration. This pattern was also observed in a semi-arid,
shallow soil (~30 cm) shortgrass habitat in Colorado, where soil moisture was short-
lived in irrigated plots at 0–5 cm soil depth (Aguilera & Lauenroth, 1995).
Consequently, the survival of Bouteloua gracilis (a perennial bunchgrass) was higher in
bare plots compared to plots where seedlings were competing with plant biomass
during the first year of establishment. Therefore, weekly watering in these bare
microhabitats probably ameliorated the combination of shallow and compacted soil
where moisture is short-lived which was beneficial for survival of seedlings.
Chapter 3
75
Weekly watering over summer may have been beneficial in two ways to reduce stress
for planted seedlings, particularly those in unfavourable conditions (i.e. planted with
bare roots and/or in shallow soil of bare plots). Firstly, the effect of increasing soil
moisture over the dry months would allow seedlings to grow larger root systems to
alleviate drought stress. Furthermore, there was a consistent correlation between
weekly watering and a reduction in soil temperature around B. ionthocarpa seedlings
during the warmest part of the day. Soil temperatures reached up to 50 °C in some
plots in the middle of summer. High soil temperatures have the potential to affect
seedlings by increasing the surrounding air temperature which would increase
evaporative demand and/or inducing tissue damage by direct contact (Kolb &
Robberecht, 1996). Therefore, a regular reduction in soil temperature through weekly
watering may have reduced stress in seedlings.
B) Tall heath plots
Survival in the tall heath plots was influenced by pre-translocation substrate. Seedlings
planted with potting mix generally had a higher survival in the tall heath plots,
however, this microhabitat was unfavourable for seedlings planted with bare roots.
The upper soil in tall heath plots was characterised by a thick layer of litter that was
relatively dry and loose during the late autumn planting. The bare roots of seedlings
grown in sand were planted directly into this loose litter. Therefore, they are likely to
have suffered as a consequence of inferior root to soil contact and thus less access to
water and nutrients. Interestingly, the weekly watered seedlings had considerably
higher two year survival rates than the monthly watered ones, suggesting that addition
of water did alleviate drought stress for those seedlings planted with bare roots in this
microhabitat. In contrast, seedlings planted with the potting mix were likely to have
grown roots into the deeper soil layers rather than the litter. This may explain why
watering frequency did not affect survival of these seedlings.
In terms of seedling growth, tall heath was unfavourable regardless of planting season
and substrate. Throughout the two years, seedlings in these plots were significantly
Chapter 3
76
smaller than seedlings in bare and low heath plots. This result was most evident for
seedlings grown in sand, which again reflects the slow growth of these seedlings
planted with bare roots in this microhabitat. The overall smaller size of seedlings in the
tall heath plots could be due to a combination of lower radiation, waterlogging and
thick litter. These plots were surrounded by taller and dense vegetation which could
have reduced amount of daily radiation used for seedling growth. Hocking (2011)
compared a range of environmental variables among the microhabitats in the B.
ionthocarpa translocation site, and found that there were no significant differences in
soil nutrients between microbabitat types (Hocking, 2011). However, tall heath plots
generally had higher clay contents at soil depths below ten centimetres (pers. obs.)
This clay layer may have contributed to slower seedling growth by becoming
waterlogged in the winter months.
Seedlings in tall heath plots were also more likely to turn chlorotic even when water
was plentiful in winter compared to the healthy, green seedlings in the other
microhabitat types. Seedling density remained constant across all plot treatments, and
therefore competition between seedlings did not contribute to chlorosis in this study.
Therefore, chlorosis appeared to be a sign of drought stress associated with summer
drought, but unlike seedlings in the other habitats, the seedlings in the tall heath
habitats did not recover during winter. The thick litter layer that characterised the tall
heath plots may provide an explanation for these observations. Litter can sometimes
retain moisture, but in the present study relative soil moisture in the top six cm was
generally lower in tall heath plots compared to the low and bare plots. Litter can
negatively affect seedling survival and growth due to its hydrophobic nature and
reduces the soil pH (Facelli & Pickett, 1991). Litter can also produce phytotoxins which
can reduce the germination and growth of surrounding seedlings (Rice, 1979). Muller
et al. (1968) demonstrated that regular fire in the Californian chaparral was a
mechanism to degrade toxins produced by the dominant shrub Adenostoma
fasciculatum which allowed herbs to grow. Indeed, Hocking (2011) found a lack of
weed growth in the tall heath plots (e.g. species such as Hypochaeris spp. Aira spp.,
Briza spp. and Romulea rosea), suggesting possible toxic effects of the deep litter layer
on generalist germination and establishment.
Chapter 3
77
C) Low heath plots
Weekly water supplementation in the low heath plots had a minimal effect on the
survival of potting mix seedlings but was once again beneficial for seedlings planted
with bare roots. Interestingly, taller regrowth around seedlings in low heath plots
significantly improved the chance of survival after two years. This result suggests a
facilitative effect of surrounding vegetation, presumably by protecting seedlings from
herbivory (García & Ramón Obeso, 2003), excessive heat or irradiance (Turner et al.,
1966), or by creating a moister soil environment through hydraulic lift (Gómez-Aparicio
et al., 2004; Padilla & Pugnaire, 2006).
This survival result, in combination with high growth rates in the low heath plots
regardless of watering frequency, suggest that this microhabitat type is a safe site for
early establishment of B. ionthocarpa. Low heath plots were also the most similar to
native range of this species in terms of vegetation cover and species composition
(Hocking, 2011). Past translocations have also used ecological similarity between
natural habitats and host sites to determine safe sites for translocation. While some
studies have found that transplants have higher success in habitats similar to their
natural habitat (Lawrence & Kaye, 2009; Maschinski et al., 2004; Montalvo & Ellstrand,
2000; Raabová et al., 2007), others have found that transplants can establish well in
dissimilar habitats (Maschinski et al., 2012; Roncal et al., 2012). While species may
have variable safe site preferences, experimental manipulations can assist with
identifying optimal microhabitats for future translocations.
Management Implications
This translocation study has highlighted some key factors that can improve seedling
survival and growth of species (particularly slow growing ones) in Mediterranean-
climate regions.
Chapter 3
78
Germinating seeds in late summer and allowing them to mature in warm
conditions might be the best approach to ensure the production of large
seedlings for translocation.
Seedlings with roots less than 5 cm long at planting had a very low (~10%)
chance of survival when roots were exposed to site soil. It is likely that growing
seedlings in pots of at least 15 cm in depth would allow for adequate root
growth, which would improve early establishment.
Planting early in winter is usually desirable to maximise the window for
establishment prior to the onset of the summer drought. However, some years
will have below-average rainfall and winter rains may arrive late. This was the
case in 2010 when the B. ionthocarpa seedlings were planted, but despite this,
survival did not decrease until the beginning of summer. Given the uncertainty
in predicting the onset of the winter rains, it might be beneficial to irrigate over
winter but this could be less frequent than during summer due to a much lower
evaporative demand.
As indicated here and in Chapter 2, growing seedlings in potting mix and
planting a root/soil plug into the field soil reduced the chance of root damage,
and so increased the chance of establishment. Planting seedlings in this manner
also requires less time compared to washing the sand off roots. This point is
particularly important when establishing large-scale translocations sometimes
with many volunteers.
Planting seedlings in a range of microhabitats that are similar and dissimilar to
their natural range can help determine safe site requirements for a species.
Clearing dense vegetation from a habitat with likely higher water supply before
planting is not always beneficial. Other amendments, such as removing some of
the litter layer to reduce the chance of allelopathic effects, may be required to
control the indirect effects of existing vegetation.
Chapter 3
79
When planting seedlings in relatively open, shallow-soil habitats regrowth that
appears in response to the disturbance can sometimes be beneficial for the
survival and growth of target species. These plants can have a “nurse effect”,
presumably by lowering soil temperatures, decreasing exposure to wind and
direct sunlight as well as increasing soil moisture.
The effectiveness of watering on seedling survival depended on many aspects
(i.e., microhabitat, seedlings size at planting, planting method) and therefore
an adaptive watering strategy might be a more effective. This would probably
require more frequent monitoring visits or sophisticated (but more costly)
equipment. For example, rain gauges or soil moisture probes could control
watering through a computer system.
Chapter 4
80
Chapter 4
The study of post-fire seedling establishment can
guide decisions for conservation management in
Mediterranean-climate communities
ABSTRACT
Incorporating ecological concepts into plant translocation programs has the potential to
inform practice to improve the survival and growth of planted seedlings. Understanding
microhabitat preferences or safe sites for seedling establishment in nature, outside of the
translocation context, is one way to do this. This study followed the fate of newly
established Acacia and Banksia seedlings over the first summer in three post-fire
environments within southwest Australia. Abiotic and biotic variables were measured
around individual seedlings and these environmental variables were correlated with seedling
survival and growth during summer. Herbivory was found to be the main factor driving
differences in seedling survival, and Acacia species were preferred by herbivores over
Banksia species at one site. Other environmental factors such as microtopography and the
presence of ash beds were also significant factors explaining differences in seedling survival
and growth. This study has confirmed the need for herbivore exclusion when establishing
translocation programs. Some relatively novel solutions could be implemented to improve
seedling establishment in Mediterranean-climate regions. These involve planting seedlings
close to dead branches/burnt vegetation which can provide protection from herbivores as
well as planting seedlings in ash beds.
INTRODUCTION
With habitat fragmentation and a range of other threats intensifying across the globe
at an accelerated rate, many plant species and communities are at risk of extinction.
Threatened plant translocation programs are increasingly being utilised by land
managers to ameliorate the impacts of human disturbance and improve connectivity
between habitats, thereby reducing the risk of extinction. Projects include
conservation efforts focused on the recovery of one species at single or multiple sites
up to large-scale conservation projects involving multiple species (Young, 2000).
Independent of scale, establishing a successful translocation needs to account for a
variety of logistical and ecological factors. In particular, understanding the factors that
Chapter 4
81
influence seedling recruitment and establishment have the potential to improve
translocation programs (Méndez et al., 2008; Valladares & Gianoli, 2007).
Germinating seed ex-situ and planting seedlings into a specified site has become a
common protocol for many translocation programs. Consequently, land managers are
often concerned with microhabitat variables that inhibit seedling survival after
planting, such as soil water availability over summer droughts in Mediterranean-
climate regions. The safe site hypothesis can help guide these decisions as it predicts
that seeds will land on a heterogeneous landscape and a range of abiotic and biotic
variables will determine which seeds survive and move through to the next life stage
(Grubb, 1977). Species will have specific requirements for safe sites based on their
tolerance to biotic and abiotic conditions. Therefore, by experimentally manipulating
these conditions within translocations, land managers can characterise a safe site for a
given species.
In a comparison of seedling establishment studies across a range of ecosystems, Moles
& Westoby (2004) found that herbivory, drought and pathogen attack were the main
factors inhibiting seedling survival, while seedling-seedling competition was rarely a
cause of death. However, mature vegetation can suppress seedlings, as demonstrated
in some Mediterranean-climate regions (Callaway & Walker, 1997; del Cacho & Lloret,
2012; Lloret et al., 2005). In such cases, disturbance can be important for seedling
establishment as it reduces the density and the resource demands of mature
vegetation which subsequently improves seedling access to soil water and nutrients.
Following the fate of newly established seedlings after a disturbance such as fire is one
method to understand how seedlings respond to a lack of competition from mature
vegetation.
Southwest Australia is a Mediterranean-climate region where seedling recruitment
commonly occurs following disturbance by fire (Lamont et al. 1993; Tyler, 1995). It is
well accepted that fire drives recruitment and that post-fire drought stress is a strong
filter for seedling establishment (Méndez et al., 2008). Post-fire recruitment is thought
to be advantageous due to reduced competition for water from mature vegetation and
a short-term increase in soil nutrients (Buhk et al., 2007). However, a scorched
Chapter 4
82
landscape provides little protection from high winds and temperature and newly
germinated seedlings are often exposed to herbivores (Gillespie & Allen, 2004; Keeley
& Fotheringham, 2000).
With these factors in mind, defining a safe site for seedling establishment after fire can
be complex. Studies to date have observed that microhabitats favourable for initial
germination are not always optimal for ultimate survival. For example, Lamont et al.
(1993) found discrepancies between those microhabitats favourable for seedling
recruitment (litter patches) of Banksia and Hakea species and those favourable for
seedling survival (sand patches due to competition for water in litter) following a
wildfire in southwest Australia. Franzese et al. (2009) observed a similar phenomenon
in Argentina with Senecio bracteolatus seedlings initially being most abundant under
tussock grasses due to a “nurse effect” which lowered soil temperatures in the first
summer after fire. However, after two years, there was a higher abundance of
seedlings in gaps due to the high level of competition under the tussock grasses. Post-
fire observations in natural habitats such as these, can inform translocation efforts to
improve seedling survival and growth during the early establishment stages.
Translocations of threatened plants have been implemented in southwest Australia
since 1998 to reduce the chance of species extinction of threatened plants in the wild.
Translocations usually involve collecting seed from remaining wild populations,
germinating seeds ex-situ and planting the seedlings at a translocation site. These
seedlings are vulnerable to water stress after planting, particularly during the very hot
and dry summers typical of the local Mediterranean-climate. Therefore, water is often
supplemented to reduce the water stress of planted seedlings (Barrett, Dillon et al.,
2011). Herbivory has also been identified as a factor inhibiting seedling survival and
growth and thus it has become common practice to erect wire fences around
translocated seedlings (Monks et al., 2012). Other manipulations to improve the
establishment of planted seedlings include mulching to increase water retention of the
soil, shading for protection from radiation and wind, and mounding and ripping to
loosen soil (Monks, 2008). While not yet considered in most translocation programs to
date, creating ash beds might also be a useful approach by providing a safe site for
Chapter 4
83
seed germination and seedling survival in fire-prone ecosystems. For example, ash
beds significantly increased survival and growth of planted seedlings of Eucalyptus
gomphocephala in degraded woodland (Ruthrof et al. 2010). Understanding how ash
beds and other variables in a post-fire environment influence seedling establishment
could assist with microhabitat manipulations in future translocations.
Acacia and Banksia are iconic genera of Australia and make up a considerable
proportion of threatened plant translocations in Western Australia with 20% of the
current translocations targeting species in these genera (Coates, D, pers. comm.,
March 2013). Acacia and Banksia have quite different life history traits but both are
adapted to post-fire seedling establishment. For example, Banksia species usually have
large seeds that drop close to the parent plant after fire and produce large seedlings
with long tap roots (Hammill et al., 1998; Lamont & Groom, 1998). In contrast, Acacia
seeds are usually smaller than Banksia seeds and have a lipid rich appendage (aril) to
attract ant and birds that disperse the seeds widely (Davidson & Morton, 1984; Hallett
et al., 2011; Shea et al., 1979). Acacia seedlings are often disturbance opportunists
that have a fast height growth strategy after fire. Safe sites for establishment are
therefore likely to vary between the genera. For example, Banksia seedlings could
have higher survival and growth in microhabitats close to dead parent plants while
Acacia seedlings may be found in highly variable microhabitats. As discussed in
Chapter 1, translocations of some threatened Acacia and Banksia species have been
successful, while others have had variable success (Barrett et al., 2011; Monks &
Coates, 2002). Therefore, implementing post-fire characteristics into translocations of
Acacia and Banksia have the potential to improve initial seedling survival.
To determine seedling recruitment after fire, the survival rates of seedlings of Acacia
and Banksia species were monitored at three sites in the south coast region of
southwest Australia. The main aim was to understand how abiotic and biotic variables
interact at the microhabitat level to create safe sites for seedling establishment in
post-fire environments. Because summer drought is a strong filter, it was hypothesised
that improved seedling survival and growth would be associated with abiotic variables
that increase soil moisture such as ash beds, litter and concave surfaces.
Chapter 4
84
METHODS
Site Descriptions
Three sites located in the south coast region of Western Australia were chosen for
study. All sites had been subject to prescribed burns by the Department of Parks and
Wildlife in June/July 2011: Torndirrup National Park (35°6'23.3"S 117°55'27.3"E),
Waychinicup National Park (34°52'55.9"S 118°19'34"E) and the Stirling Range National
Park (34°18'50.8"S 117°59'23"E). Autumn fire promotes seedling regeneration in this
region. These sites are characterised by dry summers and cool wet winters.
Torndirrup National Park is situated on the peninsula just south of Albany township.
The climate is cool and windy, and strongly influenced by maritime conditions. The
burnt area was slightly sloping (10°) in a northerly direction and soil on site was loose
alkaline sand with granite bedrock (Sandiford & Barrett, 2010). The study site was
approximately 40 m × 35 m in area and dominated by coastal heath. Ground cover to
30 cm consisted of rushes, sedges and grasses with a dense overstorey to three metres
consisting of woody perennials in the genera Acacia, Agonis, Banksia and Hakea.
Waychinicup National Park is located 65 km northeast of Albany. The burnt site was
adjacent to the road to Waychinicup Inlet on a 10° ENE facing slope. The study area
was approximately 30 m × 30 m. This site occurs on poorly draining soils with
underlying laterite or spongolite (Sandiford & Barrett, 2010). Vegetation was
characterised by a ground cover of sedges and rushes (to 30 cm), a dense understorey
(to one metre) dominated by woody perennials in the genera Allocasuarina, Banksia,
Hakea, Kingia and Taxandria. The overstorey was over two metres and dominated by
Hakea and Eucalyptus species.
Stirling Range National Park is 80 km north of Albany in a mountainous region. The
burnt area was on the northern boundary of the National Park with farms to the west
and north. The study site was in the centre of the burnt area with an area of
approximately 65 m × 65 m. The soil was sandy and shallow (4–5 cm) with exposed
rocks. The site was characterised by sparse mixed mallee and scrub heath vegetation
with low-lying grasses up to 30 cm, Banksia, Callothamnus, Hakea, Melaleuca and
Chapter 4
85
Xanthorrhoea species up to one metre, and an open overstorey dominated by
Eucalyptus x tetragona at 4–5 m in height.
The sites were characterised by a burnt ground cover and understorey (some larger
overstorey Eucalyptus remained alive at Stirling Range and Waychinicup). There was a
high density of charcoal limbs of some understorey species at the Waychinicup site.
The soil was a mosaic of ash beds and bare sandy soil as well as some woody litter as a
result of the fires. These observations suggest that the fires were relatively uniform
among sites.
Data Collection
Seedlings of a total of three Banksia and three Acacia species were surveyed at the
three sites in October 2011 (i.e. spring; Table 1). These species were easily identified at
the seedling stage by the distinctive shape of their first leaf (Banksia) and phyllode
(Acacia). The survey was constrained by local sites that had been burnt and the Acacia
and Banksia species that germinated after fire. Consequently, different species were
tagged at the three sites and the Stirling site was the only site to have both genera
represented in the dataset.
A total of 179 plots (150 × 150 mm) were placed within two distinctive microhabitat
types (ash and non-ash) at the three sites (Stirling n = 118, Torndirrup n = 30,
Waychinicup = 21). Ash microhabitats were distinguished from non-ash microhabitats
by the black powdery remains of burnt vegetation. At each site, I tagged up to 74
seedlings of each species, equally divided over the two microhabitats. The number of
tagged seedlings per plot ranged from one to ten seedlings. The plots were centred on
a target seedling and seedlings were tagged with a small PVC ring (Figure 1). The
measurements at the seedling level were height (mm), microtopography around each
seedling (concave, convex or flat) (Jumpponen et al., 1999) and distance to and height
of nearest seedling neighbour. I could not follow the fate of 26 Banksia mucronulata
seedlings at Waychinicup because the tags were removed by wildlife.
Chapter 4
86
The percentage cover of rocks, ash (or incorporated ash mixed with soil), animal scats
(rabbit scats at the Stirling site were small and round while kangaroo scats were larger
and round/ovoid), leaf and twig litter (defined as dead, unattached and unburnt
matter) and living plant cover was estimated at the plot level. The distance, identity
and height of the nearest surviving adult (usually a resprouting woody species) was
measured from the centre of each plot. Soil moisture was measured from the centre of
each plot to a depth of ten centimetres with a moisture probe (MPM-160-B, ICT
International, NSW Australia).
Table 1 – Species, average seed size, study site and number of seedlings tagged for survey. In
October 2011, 74 seedlings of each species were tagged but some tags were misplaced or
removed by wildlife between tagging and April 2012. Seed sizes were measured by Lewis, B
(pers. comm. 29th November 2013) and own seed collection.
Species Average seed
size (mg) Location
Number of Seedlings
Banksia blechnifolia 85 Stirling Ranges 73 Banksia tenuis 10 Stirling Ranges 71 Acacia chrysocephala 3.5 Stirling Ranges 55 Acacia saligna 9 Stirling Ranges 57 Acacia littorea 4.5 Torndirrup 60 Banksia mucronulata 11 Waychinnicup 34
In March 2012, after the summer drought, the mortality and height of live tagged
seedlings were recorded. Extensive herbivory was observed at some sites; seedlings
eaten by herbivores were easily identified as those with no leaves and only a small
amount of stem (Figure 1).
Chapter 4
87
Figure 1 – Tagged seedlings eaten to the soil surface by herbivores (left) and uneaten seedlings
(right).
Data Analysis
The statistical package ‘FactomineR’ in ‘R’ was used for initial exploration of data (R
Core Team, 2013). Plot level measurements (ash/non-ash, % rock, % litter, % living
matter, % scat, distance to nearest woody resprouters, height of nearest woody
resprouters, density of tagged seedlings and soil moisture) and site-specific climate
variables (number of days without rain Oct 2011 – Mar 2012, total rainfall between fire
and seedling tagging, and total rainfall Oct 2011 – Mar 2012) were used for a principle
component analysis (PCA) to compare the environment of the three sites based on a
correlation matrix. ‘Plot’ and ‘site’ were used as supplementary variables to assist with
interpretation but not used in the calculation of distances among environmental
variables. To account for missing values for some environmental variables, a
regularised iterative PCA algorithm was used to avoid problems with over-fitting as
demonstrated by Josse et al. (2012).
Classification and regression trees were created in the ‘tree’ package in ‘R’ to identify
interactions between survival and growth of seedlings and microhabitat variables to
guide the structure of mixed models. All variables were included in initial trees and
subsequently 50 sets of 10 fold cross validations were used to find the optimal tree
that explained the most variation with the least variable and the lowest estimated
error (results not shown; De'ath & Fabricius, 2000; Ripley, 2013).
Chapter 4
88
The factors ‘species’ and ‘herbivory’ were highly correlated in the classification tree
and moderate collinearity was confirmed using a calculation of kappa (the ratio of the
largest to smallest non zero singular value in a matrix). Residualisation was used to
reduce collinearity between these two variables, herbivory was regressed with species
to create a new variable which was the difference between herbivory rate and
herbivory predicted by species (Hofmeister, 2010). The packages ‘nlme’ and ‘lmer’
were then used in ‘R’ to create multivariate mixed models in order to analyse survival
and growth of tagged seedlings. This residual variable and all other variables were
used in a generalised linear mixed model (GLMM) with a binomial distribution with
plot as a random factor and mortality as a binary variable (1 = dead, 0 = alive) (R Core
Team, 2013). Some continuous variables (e.g. distance to nearest seedling neighbour,
distance to woody resprouters) were centred to reduce collinearity. Initially, a full
model containing all measured variables was constructed, after which a step-wise
exclusion procedure was used to determine the best fitting model (Zuur et al., 2007).
Interactions between variables were added using the PCA as a guide. Data likelihood
was estimated using the Laplace approximation. Models were compared using a chi-
squared distribution and the model with the lowest AIC was selected if differences
between models were statistically significant. When the optimal model was found,
each variable was individually dropped and the change in deviance was used to
determine the contribution of each variable to the optimal model (Mellin et al., 2010).
A similar approach was used to compare changes in seedling height with the
environmental variables. Plot was included as a random factor (intercept) and change
in height (i.e., final height – initial height/ initial height) was included as the response
variable. Due to extensive herbivory, both A. saligna and A. chrysocephala only had a
small number of living seedlings by March 2012 (n = 4 and 5 respectively). Therefore,
these species were excluded from the mixed models of seedling growth. The maximal
model contained all explanatory biotic and abiotic variables. Model residuals were
compared with each variable in the model to identify non-normality in the data. A
number of the continuous variables were clustered around zero and were therefore
centred to reduce the multicollinearity effects. Furthermore, the variance among
species and plot location (ash or non-ash) caused non-normality in the data (due to
Chapter 4
89
differences in mortality) so the error terms (slopes) were allowed to vary among
species. Two continuous variables: height of woody resprouters and distance to
nearest neighbour showed highly heterogeneous residuals and a power function was
used to improve their distribution (Zuur et al., 2007). From the maximal model,
variables were subtracted in a step-wise fashion and the AIC was compared with a
likelihood ratio test. A maximum likelihood (ML) model framework was used during
step-wise comparison of models to find the parameters that maximised the probability
of a response. The optimal models had the lowest AIC at the 5% significance level. A
restricted maximum likelihood (REML) framework was used to report parameter
estimates of the optimal model as this method accounts for the random plot factor
and is less biased compared to a ML framework (Bolker et al., 2009).
RESULTS
Climatic and environmental conditions at the site level
All sites experienced above-average rainfall in spring, particularly the Stirling Range
site that received 2.4 times more rainfall than the long term average (Table 2).
Subsequently, all sites received lower than average rainfall over summer (Stirling
Range, Torndirrup and Waychinicup received 42, 38 and 36% less rainfall respectively
compared with long term). The low rainfall in spring and summer differentiated the
plots in Stirling Range from those at Waychinicup and Torndirrup (PC1; Figure 2).
Overall, the Torndirrup site had most rain and Stirling site the least rain during the
period of study.
PC1 also highlighted some differences between plot variables across sites. Torndirrup
and Waychinicup sites had higher soil moisture, % litter and taller woody resprouters
(Figure 2; Table 2). In contrast, Stirling Range was differentiated by a more sparse
vegetation cover indicated by a negative correlation with distance to woody
resprouters on PC1. PC2 explained 12.3% of plot variation which demonstrated that
Chapter 4
90
the Stirling Range site had a greater range of tagged seedling densities and % green in
plots and these variables were negatively correlated with location in ash or non-ash
microhabitats (Figure 2; Table 2).
Table 2 – Rainfall (mm) characteristics of the three experimental sites before and during the
survey period compared with long term rainfall (1958 – 2010, 1968 – 2010 and 2001 – 2010 for
Stirling Range, Torndirrup and Waychinicup respectively). Pre-sampling (Aug – Oct 2011), late
spring (Nov – Dec 2011) and summer rainfall (Jan 2011 – Mar 2012) and the longest number of
days without measurable quantities of rainfall (Oct 2011– Mar 2012) was obtained from the
nearest meteorological station. Rainfall data was sourced from the North Stirling climate
station which was 10.5 km south Stirling Range site; Little Grove climate station was 7.4 km
south of Torndirrup site and Cheyne Beach climate station was 6.4 km north of Waychinicup
(Bureau of Meterology, 2013).
Site Pre-
sampling rainfall
Late spring rainfall
Long term average
rainfall in late spring
Summer rainfall
Long term average
rainfall in summer
Number of consecutive
days without rain (October
2011 to March 2012)
Stirling Range 137 180 74 20 48 17 Torndirrup 274 197 159 31 81 20 Waychinicup 206 185 133 27 74 22
Chapter 4
91
(A)
(B)
Figure 2 – The first two axes of the PCA used to differentiate the three post-fire sites (Stirling,
Torndirrup and Waychinicup) by climate and microhabitat. (A) Site climatic variables (drought
= number of consecutive days without rain, rainspring = total rainfall between Oct 2011 – Dec
2011, rainsummer = total rainfall between Dec 2011 – Mar 2012) and plot variables (% rock, %
plant matter, % litter, ash/non-ash, % scat, density of tagged seedlings per plot, height of
nearest woody neighbour and distance to nearest woody neighbour) were included in the PCA.
Plot and site were supplementary variables (not included in the calculation of the PCA) which
were used to display the spread of individual plots in (B). The numbers in blue denote the plots
at the Stirling Range site, Waychinicup plots are indicated by red and Torndirrup plots are in
grey.
PCA
2 (1
2.3%
)
PCA 1 (27.8%)
Chapter 4
92
Seedling Survival
Survival rates significantly varied among species (Figure 3; Table 3). Herbivory caused
high mortality of seedlings of both Acacia species at the Stirling Range site, with 97%
and 100% of mortality in A. saligna and A. chrysocephala, respectively, due to
herbivory (Figure 3). In plots that contained both Acacia and Banksia species (n = 11)
there was evidence for preferential herbivory on Acacia species. In 63% of these plots
Acacia seedlings had all been eaten while the Banksia seedlings remained untouched.
In the most extreme case, six of the Acacia seedlings were eaten while the four
Banksia seedlings in the same plot remained intact.
There was a significant interaction between herbivory and distance to nearest woody
resprouter (Table 3). Acacia saligna seedlings that were not damaged by herbivores
were significantly closer to woody resprouters (ave. 427 mm ± 132) compared with
those seedlings that had been eaten (ave. 586 mm ± 69).
Figure 3 – Percentage mortality of tagged seedlings for each species which were
differentiated by cause of death (herbivory or unknown) between October 2011 and
March 2012. Stirling Range site: A. chrysocephala (n = 55), A. saligna (n = 57), B.
blechnifolia (n = 73), and B. tenuis (n = 71). Torndirrup site: A. littorea (n = 50) and
Waychinicup: B. mucronulata (n = 34).
0 10 20 30 40 50 60 70 80 90
100
A. chrysocephala A. littorea A. saligna B. blechnifolia B. mucronulata B. tenuis
Tot
al m
orta
lity
(%)
Eaten Unknown
Study species
Chapter 4
93
Table 3 – The optimal mixed model for seedling survival over their first summer as based on measured variables at the seedling (species, microtopography, distance to nearest seedling neighbour) and plot level (ash/non-ash, % rock, % litter, % living matter, % scat, distance to nearest woody resprouters, height of nearest woody resprouters, density of tagged seedlings and soil moisture). Plot was used as a random variable in the model. A maximal model was created with all measured variables and step-wise deletion was used to find the optimal model. An initial classification tree was used to guide the addition of interaction terms into the model. Percent deviance is used to determine the amount of variation explained by each of the main variables in the optimal model. For the categorical variable ‘microtopography’, ‘flat’ was used as a reference level. The grand mean of ‘species’ was used as a reference level for each species level. Bold p-values indicate significance ≤ 0.05.
Variables included in the optimal model
% Deviance explained in optimal
model
P-value
Seedling Level
All species 18.7%
Acacia littorea 0.005
Acacia saligna <0.001
Banksia blechnifolia 0.02
Banksia mucronulata 0.07
Banksia tenuis 0.26
Herbivory 32.4% 0.002
Distance to nearest seedling neighbor 9.3% 0.96
All microtopographies 21.3%
Convex topography 0.01
Concave topography 0.25
Plot Level
Soil moisture 10.9% 0.17
Distance to woody resprouter 7.9% 0.54
Interactions
Herbivory × distance to woody resprouters 0.01
Distance to nearest seedling neighbour × convex topography 0.19
Distance to nearest seedling neighbour × concave topography 0.02
Chapter 4
94
Survival also varied based on microtopography. For three of the six species there was
evidence that seedlings on convex surfaces had significantly lower survival than
seedlings on flat and concave surfaces (Table 3). There was also a significant
relationship between microtopography and distance to nearest neighbour for the
Banksia species. That is, dead Banksia seedlings located on concave surfaces were
significantly further away from their nearest neighbour (ave. 60 ± 13 cm) compared
alive Banksia seedlings on the same surface (ave. 34 ± 7 cm; Table 3).
Seedling Growth
Species was the most significant factor explaining changes in height (Figure 4). Banksia
species showed a modest increase in height during the sampling period, while A.
littorea and A. chrysocephala seedlings had significantly greater growth.
Figure 4 – Percentage height increase ± SE of tagged seedlings during their first summer since
establishment (Oct 2011 – March 2012). Acacia chrysocephala (n = 4), A. saligna (n = 4), B.
blechnifolia (n = 61) and B. tenuis (n = 52), were tagged at Stirling Range site, B. mucronulata (n
= 27) and A. littorea (n = 54) were tagged at the Waychinicup and Torndirrup sites respectively.
0
100
200
300
400
500
600
700
A.chrysocephala A. littorea A. saligna B. blechnifolia B. mucronulata B. tenuis
Incr
ease
in h
eigh
t (%
)
Study species
Chapter 4
95
Differences in seedling height within a species were correlated with different
microhabitat variables. Firstly, increases in height of Acacia littorea seedlings were
significantly correlated to the height of the nearest woody resprouters (Table 4).
Furthermore, both Banksia species at Stirling Range (B. blechnifolia and B. tenuis) were
significantly taller in ash compared to non-ash environments, but this was not true for
A. littorea and B. mucronulata seedlings in Waychinicup and Torndirrup, respectively
(Figure 5). Finally, increasing soil moisture had a negative relationship on the heights
of B. mucronulata and B. blechnifolia seedlings.
Table 4 – Summary of optimal mixed model with height increase of seedlings (%) as related to
microhabitat variables collected from the plot level (ash/non-ash, % rock, % litter, % living
matter, % scat, distance to nearest woody resprouters, height of nearest woody resprouters,
density of tagged seedlings and soil moisture) and seedling level (species, microtopography
and distance to nearest seedling neighbour). Some variables, such as soil moisture, were
initially clustered around zero and centred before being added into the models. Plot was used
as a random variable. A maximal model was first created with all variables and step-wise
deletion was used to determine the optimal model with the lowest AIC. For the categorical
variables ‘location’ and ‘topography’, ‘non-ash’ and ‘flat’ were used as reference variables. The
grand mean of all species was used as the reference level for each species. Bold values indicate
p ≤ 0.05.
Variable Degrees of freedom
p-value
Plot Level Non-ash location 87 0.03 Soil moisture (centred) 87 0.007
Seedling Level
Acacia littorea 96 <0.001
Banksia blechnifolia 96 0.01 Banksia mucronulata 96 0.29
Interactions
Soil moisture × Acacia littorea 96 0.08 Soil moisture × Banksia blechnifolia 96 0.02 Soil moisture × Banksia mucronulata 96 0.02
Chapter 4
96
Figure 5 – Average percentage increase ± SE in seedling height between October 2011 and
March 2012 as related to microhabitats (ash or non-ash) for A. littorea seedlings monitored at
Torndirrup (N = 38 ash; 17 non-ash), seedlings of B. blechnifolia (N = 36 ash; 25 non-ash) and B.
tenuis (n = 25 ash; 27 non-ash) at Stirling Range and B. mucronulata at Waychinicup (n = 12
ash; 15 non-ash). Acacia chrysocephala and A. saligna were not included due to the small
number of surviving seedlings. Asterisks indicate significant height differences within a species
(p<0.05).
DISCUSSION
This study has highlighted the complex nature of safe sites that can vary among species
and sites. Highly variable safe sites have been observed in natural recruitment
(Eldridge et al., 1991) and manipulated systems (Fowler, 1988). Herbivory was a
significant variable controlling the composition of Acacia and Banksia seedlings at the
Stirling Range site and there was some evidence of preferential feeding by large
herbivores. Microtopography was the main abiotic variable that influenced seedling
survival, probably due to the movement of water and/or nutrients down convex
surfaces and pooling in flat/ concave ones. Increases in seedling height were largely
based on site and species. Acacia littorea seedlings grew rapidly at the wettest site
Torndirrup to become the dominant mid-storey vegetation (as observed in the
neighbouring unburnt vegetation). The growth of Banksia species at the drier Stirling
Range sites was enhanced in the presence of ash beds. This study demonstrates the
0
100
200
300
400
500
600
A. littorea B. blechnifolia B. mucronulata B. tenuis
Incr
ease
in h
eigh
t (%
)
Study species
Ash
Non-ash *
Chapter 4
97
importance of incorporating different microhabitats into translocation projects and
contributes to the understanding of ecology in post-fire environments.
Survival
Herbivory was a major biotic factor driving significant differences in seedling survival.
Herbivory was prevalent among the Acacia species (A. saligna and A. chrysocephala) at
the Stirling Range site, while apparently absent among A. littorea seedlings at
Torndirrup, and insignificant in all Banksia species. While some studies have found
herbivory to be negligible (Bradstock, 1991; Caughley et al., 1985; Hanley & Lamont,
2001), others have shown it to be intense in various post-fire environments in
Mediterranean-climate regions (Cohn & Bradstock, 2000; Davis et al., 1989; Parsons et
al., 2007; Tyler, 1995). There are some possible explanations for this pattern. Grazing
was likely due primarily to rabbits or kangaroos, as there was evidence of fresh rabbit
and kangaroo droppings at the Stirling Range site. Unlike Torndirrup, the Stirling Range
site was surrounded by farmland on three sides of the reserve and a road on the
fourth. The farmland would likely support large populations of rabbits and kangaroos
(Hobbs, 2001). In contrast, Torndirrup was within the boundaries of a national park
and consequently remote from main roads and farmland. It was also differentiated
from the Stirling Range by a wetter climate and more dense mature vegetation, which
may have reduced the density of herbivores. Thus, in the more open and drier habitat
of the Stirling range, forage may have been more limiting and herbivores more
attracted to the fresh new growth after the fire (e.g. Allred et al., 2011; Caughley et al.,
1985). Indeed, Caughley et al. (1985) demonstrated a lack of response by kangaroos
after fire in a mallee woodland in NSW, which was hypothesised to be due to plentiful
rains and abundant surrounding vegetation. The lack of herbivore evidence at
Torndirrup could also be due to species-specific differences in palatability among
Acacia species including toxic compounds or phyllode morphology. Further research
would assist with determining if site or species characteristics were driving differences
in herbivory.
Chapter 4
98
Herbivores also displayed a preference for Acacia over Banksia species at the Stirling
Range site. Selective herbivory has been observed in kangaroos (Parsons et al., 2006;
Rafferty et al., 2005) where Hakea seedlings with a lower phenolic content were
targeted regardless of their physical defence (i.e. leaves reduced to spines). Jones et al.
(2003) found kangaroos were more likely to feed on Acacia (Acacia saligna and Acacia
iteaphylla) compared with Banksia (Banksia grandis and Banksia menziesii) and
suggested that this result was due to the phenolic components of the latter. Hanley &
Lamont (2001) found a significant correlation between herbivory attack by
invertebrates and percentage phenolics of twelve species of Proteceae from Western
Australia including Banksia mucronulata. Banksia mucronulata had intermediate
phenolics content (1.6%) compared to the other species sampled (highest in Hakea
cucullata ~3.5%) and an intermediate level of herbivory. Preferential herbivory could
also be attributed to differences in nutrition. For example, Shea et al. (1979) observed
that fast-growing Acacia species were a preferred source of nutrition for marsupials.
Thus, B. mucronulata seedlings and seedlings of the other Banksia species were not a
target for herbivores in this study potentially due to their phenolics content and/ or
lower nutritional content.
Microtopography was another significant factor influencing survival of tagged
seedlings. Some studies have found higher survival on convex surfaces in regions with
alternate seasons of flood and drought, due to less fluctuating soil water (Barberis et
al., 2002), while others have found no effect (Bannister et al., 2013; Maestre et al.,
2003) or higher survival on concave surfaces (Battaglia & Reid, 1993; Whisenant et al.,
1995). In the present study, Acacia seedlings had a significantly lower survival on
convex surfaces compared with those growing on flat and concave surfaces. This was
partly related to the higher herbivory on convex surfaces. However, the total number
of Acacia littorea seedlings that died (n = 6) were not subject to herbivory and were all
located on convex surfaces. This suggests that seedlings are negatively affected by
convex surfaces, which could be related to their lower moisture retention as rainfall
would likely run off these surfaces and pool in concave ones. Indeed, Davis et al.
(1989) found a positive correlation between gaps (associated with microtopographic
Chapter 4
99
depressions) and survival of coastal chaparral species. They attributed this survival to
higher soil moisture and nutrient concentrations in the depressions.
Growth
Growth of tagged seedlings varied across sites and species and the effect of
microhabitat variables were species-specific. Acacia littorea and A. chrysocephala
grew up to 600% taller over their first summer compared to Acacia saligna and the
Banksia species. These differences were predominately due to varied growth
strategies where Acacia seedlings often have a taller growth form for a given stem
width while Banksia seedlings often have a wide growth form (C. Allen, pers. obs).
Acacia is known to use a fast-growing strategy following fire, which has been
attributed to their ability to fix nitrogen (Barrow, 1977; Witkowski, 1991). Acacia
littorea was the dominant shrub species at Torndirrup which was the wettest site. The
PCA showed that Torndirrup had a higher litter cover and a closer distance to nearest
woody resprouters which correlated with a dense 1–2 metre high shrub layer in the
unburnt patches. The combination of high water availability and dense vegetation
cover could mean that A. littorea used a fast-growing strategy to access light before
the surrounding seedlings and resprouters could dominate. In some cases these tagged
seedlings were nearly covered by resprouters at the Torndirrup site by March 2012 (C.
Allen, pers. obs.) This “race” for space and light was also observed in a range of
sclerophyllous species in a post-fire site in southeast Australia, particularly those
species with shorter mature heights (Falster & Westoby, 2005). When comparing
growth of the two Acacia species at Stirling Range, A. chrysocephala was significantly
taller than A. saligna. This result was unexpected given that A. saligna can reach up to
nine metres at maturity while A. chrysocephala is a small sub-shrub to 0.5 m. The small
sample size of both species due to herbivory probably contributed to the high variation
in growth measures.
Banksia blechnifolia at the Stirling site grew significantly taller than B. tenuis at the
same site and B. mucronulata at Waychinicup. Differences in mature growth form did
Chapter 4
100
not explain these height differences as both Banksia species at the Stirling site have
prostrate growth form at maturity (0.7–1 metre tall) while B. mucronulata is a taller
shrub to 2.5 metres. However, B. blechnifolia had ~85% larger seeds compared to the
other Banksia species. Seed size has been significantly correlated to seedling size in
woody perennial species (Hallett et al., 2011; Jurado & Westoby, 1992; Lamont &
Groom, 2013; Lamont & Groom, 1998). Therefore, some of the variation in growth
could be attributed the large differences in seed size.
At the Stirling Range site, the heights of both Banksia species, particularly B.
blechnifolia, were significantly taller in ash patches. Similar “ash bed effects” have
been reported in other studies with seedling growth being stimulated in ash patches as
a result of increased nutrients, particularly phosphorus and nitrogen (Chambers &
Attiwill, 1994). For example, Rice (1993) found extractable phosphorus in ash beds was
highly correlated with species richness and total cover after fire in Californian
chaparral. Thus, Banksia seedlings at the Stirling may have grown taller in response to
increased nutrients located in the ash beds.
Implications for translocations
The results presented in this study have important implications for the establishment
of Australian native species in conservation settings. It has long been recognised that
herbivory is an important factor constraining seedling survival (Moles & Westoby,
2004). Planted seedlings could potentially be more vulnerable to herbivory compared
with recruited seedlings as they are likely to have higher leaf nutrient concentrations
and could also have softer tissue as a consequence of their early growth in the
protected nursery environment. Consequently, fencing has been used in southwest
Australia to reduce herbivore impacts on seedlings planted for translocation and
restoration projects (Monks & Coates, 2002; Yates et al., 2000). This study shows that
the effect of herbivory is highly species-specific, suggesting that it may not always be
necessary to fence translocated seedlings. For seedlings of species that are preferred
by local herbivores such as the Acacia species at the Stirling Range site in this study,
Chapter 4
101
co-planting with highly palatable fast growing species, as suggested by Parsons et al.
(2006), may alleviate some of the herbivory pressure. Their suggestion was based on
the observation that fast growing species at a mining restoration site in the Darling
Range in southwest Australia were highly preferred by herbivores. However, the
potential advantage of these “sacrificial species” may be negated by increased
herbivore attraction to the site and by potential competitive effects.
Planting seedlings close to mature vegetation (Padilla & Pugnaire, 2006; Uytvanck et
al., 2008), or under burnt branches (Castro et al., 2011) could be alternative options to
reduce the effects of herbivory on young seedlings. Using dead branches as an
herbivore deterrent was also supported by the current study as seedlings of A. saligna
were less likely to be eaten when located close to burnt resprouters. Therefore, using
mature or burnt vegetation as “nurse objects” could reduce the negative effects of
herbivory in early establishment stages, but have the potential to become a source of
competition for water and nutrients as seedlings grow larger.
Soil characteristics found to be important for seedling growth, such as fertilisation,
have been manipulated in some restoration projects (Devine et al., 2007).
Furthermore, soil ripping has been used in translocations (Monks, L pers. comm. 28th
November 2013) and restoration projects (Close & Davidson, 2003) within Australia to
loosen the soil for root growth. The current study suggests that seedlings are more
vulnerable on convex surfaces (possibly due to being more exposed to herbivory/
drying) over summer. Consequently, planting seedlings on convex surfaces should be
avoided, particularly with increasing drought. Rather, creating depressions before
planting is likely to assist with water accumulation which can improve seedling survival
and growth (Méndez et al., 2008).
Both Banksia species at the Stirling Range site were significantly taller in ash bed
microhabitats. Ruthrof et al. (2010) successfully used artificial ash beds in the
restoration of Eucalyptus gomphocephala and four understorey species in the
woodlands near Perth, Western Australia. Seedlings planted in ash beds had
significantly higher survival and growth compared to those planted in ripped beds. Ash
bed addition has not been implemented in translocations in southwest Australia but
Chapter 4
102
has the potential to improve early establishment (Barrett, S pers. comm 23rd April
2013). The current study suggests that the importance of ash beds for survival and
growth is highly species-specific and may also depend on local soil type and rainfall.
However, as no negative effects were observed, and many species in southwest
Australia regenerate after fire, it may be worthwhile to include ash beds in rare flora
translocations as a standard protocol.
Chapter 5
103
Chapter 5
Using leaf osmotic potential to predict drought
tolerance among vegetation communities across
an aridity gradient in southwest Australia
PREFACE
This chapter was initially planned to investigate the drought tolerance of threatened
and common Acacia species, one of which included Acacia awestoniana (Chapter 2).
However, this experiment had to be abandoned due to a fungal disease which caused
high mortality in the seedlings before the commencement of the drought. Instead, this
chapter uses common species in the field to explore osmotic behaviour at the plant
community level.
ABSTRACT
Drying climates associated with climate change are impacting plant communities worldwide.
Seedlings are particularly vulnerable to drought, but their physiological responses are
understudied in the field compared with the responses of mature vegetation. A range of
climatic, morphological and physiological variables need to be considered when predicting
responses of plant species and communities to a drying climate. This study aimed to
decipher the relative importance of climatic variables (rainfall, temperature), abiotic
variables (soil moisture) and leaf traits (leaf water content and leaf mass per area) in
determining one of the most important measures of drought tolerance, leaf osmotic
potential, along an aridity gradient in southwest Australia. The study was designed to
compare the leaf osmotic potential of mature plants and seedlings within and among seven
plant communities pre- and post-summer drought. Post-fire communities were targeted as
seedling recruitment predominantly occurs after fire in Mediterranean-climate regions such
as southwest Australia. Unexpectedly, late summer rainfall was higher than late spring
rainfall across the sites in the year of study. Both mature vegetation and seedlings
responded to the above-average rainfall in late summer by increasing their leaf osmotic
potentials. The wettest community had the highest community-averaged osmotic potential,
while plants in the driest community had some of the lowest values. Leaf osmotic potential
of mature vegetation was significantly correlated with leaf water content and soil moisture
at the site level, while leaf mass per area was a poor predictor of leaf osmotic potential. For
seedlings, there was a weak correlation between rainfall and leaf osmotic potential, which
suggests there are highly variable responses to drought at this stage of maturity. The inter-
and intra-specific variability in leaf osmotic potential suggests that other measures of
physiology and morphology are needed to fully understand community water relations.
These results can be used to guide further studies into water relations and are discussed in
the context of conservation of plant communities in a drying climate.
Chapter 5
104
INTRODUCTION
Climate-related mortality has increasingly been observed in mature vegetation across
the globe, including Eucalyptus species across Australia and sclerophyllous heathland
in Western Australia (Allen et al., 2010; Hnatiuk & Hopkins, 1980; Matusick et al.,
2013; McDowell et al., 2008; Poot & Veneklaas, 2013; Rice et al., 2004). Mortality due
to drought is likely to increase in southern parts of Australia, with predictions of an
increasingly dry and warmer climate and consequent increase in the severity and
frequency of drought (CSIRO, 2007; Hennessy et al., 2008; Hughes, 2003). Therefore,
understanding the levels of drought tolerance within a vegetation community would
be valuable for future management and conservation. For example, impending
collapse of a dominant species due to drought could have severe consequences for
biota dependent on these species (Anderegg et al., 2012; Slik et al., 2013).
Communities that are vulnerable to drought-related mortality may require direct
intervention to reduce other potential stressors, such as fencing to reduce grazing of
recruits or reducing competition from invasive species.
When predicting community responses to a drying climate, a species’ reproductive
success and survival of subsequent recruits is important to consider. These recruits will
ultimately determine the future composition of vegetation communities (Lloret et al.,
2004). Recruitment often occurs after a disturbance, such as fire, in Mediterranean-
climate regions. In these regions, species recruit by resprouting after fire or by
regenerating from seed. Seedlings that have regenerated from seed are particularly
vulnerable to drying due to their small root: shoot volume which limits their capacity
to acquire water (McDowell et al., 2008). Therefore, recruitment of these reseeding
species could become increasingly difficult with longer drought periods. Also, fires are
predicted to become more prevalent in an increasingly hot and dry environment.
Consequently, some reseeding species could become locally extinct as frequent fires
may disadvantage obligate seeders with long juvenile periods (Bell, 2001).
Understanding how reseeding species respond to drought at the seedling and mature
stage can provide insight into which species could be at risk of drought-related
mortality.
Chapter 5
105
Drought responses measured at the leaf level within an individual plant population in a
field environment can be scaled up to make predictions about the response of whole
communities (Bazzaz, 1996). Species have evolved a variety of physiological and
morphological responses to survive drought periods by either tolerating drought to
some degree or avoiding it. Being able to sustain a greater water potential gradient
from roots to leaves through the accumulation of cell solutes in leaf cells (i.e. by
achieving a more negative osmotic potential) is one of the major tolerance
mechanisms. This mechanism allows plants to maintain cell turgor at a lower leaf
water potential while still drawing water from the drying soil. Species from drought-
prone biomes have inherently lower osmotic potentials or develop lower potentials
through osmotic adjustment during the dry season, compared with species from
wetter biomes (Bartlett et al., 2012). Based on this correlation at the biome level, it is
predicted that leaf osmotic potentials can be used to identify levels of drought
tolerance within communities across an aridity gradient.
Grouping plant species based on their functional traits (e.g. leaf mass per unit area
LMA, seed mass and plant height) can also be used to predict their responses to
climate change. This trait-based approach enables large-scale comparisons across
communities, ecosystems or even biomes (Bartlett et al., 2012; Reich et al., 2003).
Species in drought-prone habitats such as Mediterranean-climate regions often have
relatively small sclerophyllous leaves that resist wilting due to thick cell walls
(Niinemets, 2001; Poorter et al., 2009). These dense leaves usually have a high LMA
and a low leaf water content which is often associated with low osmotic and water
potentials (Abrams et al., 1994). The highly diverse plant communities in southwest
Australia have a large number of woody perennial species that are likely well-adapted
to their seasonally dry Mediterranean climate. However, for most communities and
species it is not known to what extent functional traits and leaf osmotic potentials co-
vary over the steep local rainfall gradients.
Southwest Australia is one of several regions with a Mediterranean climate that is
predicted to become drier and warmer in the future with an increase in extreme
weather events. It is predicted that this region will experience a further 5–20%
Chapter 5
106
decrease in rainfall by 2030 (Charles, 2010). An increase in extreme events such as
longer periods without rain and more days of extreme heat (over 35°C) are also likely
(Alexander & Arblaster, 2009). Plant communities in this region are adapted to
seasonally variable soil water content. However, the combination of longer periods of
low water availability and extensive fragmentation means that species could be at risk
of climate-related extinction as they cannot migrate to more suitable climatic
conditions. For this reason, some climate models have predicted a range collapse of
flora in southwest Australia (e.g. Renton et al., 2012). Similarly, Fitzpatrick et al. (2008)
predicted a range contraction in almost 85% of the 100 Banksia species found in this
region. However, so far empirical data do not support these predictions (Standish et
al., 2012; Witkowski & Lamont, 2006). Furthermore, species distribution models do not
often consider the potential for plastic responses of species to survive drought, such as
the capacity for osmotic adjustment, or rapid evolutionary changes such as strong
selection for genotypes with inherently lower osmotic potentials.
Although low leaf osmotic potentials can be indicative of drought tolerance, perennial
woody species can also avoid drought by developing a deep root system. Past studies
investigating the physiological responses of mature plant communities to drought in
southwest Australia have emphasised the importance of deep roots for many species
to avoid drought. Dodd and Bell (1993) observed that Banksia attenuata and Banksia
menziesii on the sandplains of southwest Australia had high transpiration rates
throughout the year due to their access to groundwater at a depth of 6–7 metres.
Furthermore, four co-occurring Eucalyptus species in the southwest woodlands
showed contrasting responses to summer drought which were likely related to rooting
depth and soil texture (Poot & Veneklaas, 2013). Drought avoidance with rapid root
growth is also used by seedlings to survive drought in southwest Australia (Schütz et
al., 2002). Studies focusing on water relations of newly emerged seedlings in post-fire
environments demonstrated that some genera such as Banksia have fast growing tap
roots that access deeper water layers before the onset of summer drought (Pate et al.,
1990). Also, the higher survival rates of relatively large seedlings have been related to
their larger root system accessing deeper water sources (Hallett et al., 2011; Lamont &
Groom, 1998; Lamont et al., 1999). The importance of accessing deeper water sources
Chapter 5
107
was also evident for translocated B. ionthocarpa seedlings (Chapter 3). Shallow-rooted
woody species in southwest Australia use drought tolerance mechanisms such as tight
control of stomatal conductance and/or adjusting osmotic potential to survive soil-
water deficits (Groom, 2004). Seedlings can also use drought tolerance strategies but
often not as efficiently as their mature counterparts (Donovan & Ehleringer, 1991). For
example, generating lower osmotic potentials would attract more water, but in
seedlings would also lead to further leaf expansion, which may be counterproductive
during drought. To date, comparisons of water relations between mature plants and
seedlings of the same species are not often reported, particularly across a climatic
gradient. Comparing osmotic potentials of adults and seedlings can help to identify
whether or not drought tolerance mechanisms are consistent between life cycle
stages.
The current study was conducted to document the variation in leaf osmotic potential
within and among a range of post-fire communities across a climatic gradient in
southwest Australia. The main aim was to determine which factors (e.g. local climatic
conditions, time of year, species traits) best correlate with osmotic behaviour of the
most dominant woody reseeding species at the mature and seedling stage across an
aridity gradient. I hypothesised that leaf osmotic potential would vary across the
climatic gradient, with communities from drier sites having lower osmotic potentials
and mature vegetation having a more negative osmotic potential than seedlings,
reflecting a greater capacity for drought tolerance. Differences in leaf osmotic
potential among mature communities could also be related to other relevant plant
traits such as differences in leaf water content and leaf mass per area. Furthermore, a
combination of rooting depths from past studies, plant heights and intra-specific
comparisons were used to correlate plant morphology with osmotic potential. It was
hypothesised that mature species of low stature and presumably shallow rooting
depth would develop more negative osmotic potentials in response to drought
compared to taller canopy species with likely access to deeper soil water. The
outcomes of this study could be used to predict drought tolerance in common or
threatened species and communities.
Chapter 5
108
METHODS
Sampling sites
Seven sites along a climatic gradient spanning from Eneabba to Northcliffe in
southwest Australia were selected for study (Figure 1). These sites had been burnt in
late summer to autumn 2011 by the Department of Parks and Wildlife. Low intensity
prescribed burning is used throughout Western Australia to reduce fuel loads, protect
property and maintain a mosaic of different fire ages to maintain biodiversity. Each site
also had an adjacent unburnt area of similar vegetation communities. Unburnt sites
were usually less than 50 metres from burnt site, which allowed for sampling of
seedlings as well as mature vegetation. Most burnt areas at each site had a mosaic of
ash and non-ash patches and tree canopies were generally unaffected (Figure 2). The
seven sites were ranked on the basis of their aridity index (AI) which is used to quantify
climate conditions and correlates with water availability (Maliva & Missimer, 2012).
The AI was calculated with long term climate data as follows:
Aridity Index (AI) = mean annual rainfall
(10 mean annual maximum temperature)
Chapter 5
109
Figure 1 – Location of the seven study sites based on their aridity indexes (in brackets) across Australia. Colours used in this map will be used to distinguish study sites in subsequent figures. Note: in general rainfall in southwest Australia decreases and maximum temperature increases from the SW to the NE. Thus, sites with high AI had high mean annual rainfall and low average temperatures compared with sites with low AI.
Figure 2 – Example of burnt (left) and unburnt (right) vegetation at Motague.
Chapter 5
110
Vegetation sampling
The leaves of seedlings and adults from the most common perennial native species
were sampled at each study site to determine their osmotic potentials and measure
other leaf traits. All species were woody reseeders; i.e. species that are generally killed
by fire and rely on seed production for recruitment. Pre-summer sampling occurred in
late spring (November 2012) and post-summer sampling was undertaken in autumn
(April 2013). At all sites, seedlings of the most common woody reseeders were
identified in the burnt area (between five and ten species at each site).
Simultaneously, adults of the most common woody reseeders were sampled in the
adjacent unburnt area. Five replicate mature plants and five seedlings per species were
randomly selected. For at least two species per site leaves of both seedlings and adults
were sampled (Appendix 6). For the mature specimens, a fully-expanded, sunlit leaf
was randomly chosen and stored in a five mL vial. Multiple leaves were collected from
small leaved and needle leaved species. For seedlings, the whole plant was collected
(including roots) to confirm that each specimen was indeed a reseeder and not a
resprouter and to ensure there was adequate leaf surface area from which to express
sap in the laboratory. Mature and seedling samples were snap frozen in dry ice
immediately after collection to ensure the concentration of leaf solutes would not
change during transportation to the laboratory. Collection of specimens was stratified
over a four hour period until midday to ensure that each species was collected
throughout the morning.
Additionally, plant height, leaf water content and leaf mass per unit area (LMA) were
determined for each mature species. Plant height was recorded for each mature
individual sampled as it typically scales with belowground biomass and thus provides a
proxy for potential access to soil water (Enquist & Niklas, 2002). Species were
categorised into one of three height classes: low (0–1 m), mid (1–2.5 m) and high (2.5
m and over). Seedlings were also categorised into these height classes as based on
species height at maturity.
Chapter 5
111
Table 1 – General characteristics of the seven study sites, including aridity index, vegetation
type and the number of woody reseeder species collected from the unburnt site as categorised
by their average height (low = 0–1 metre, mid = 1–2.5 m and high= 2.5 m and over).
Site name Aridity Index
Vegetation type Height class
L M H
Wandoo National Park 12 Woodland 3 0 1 Motague State Forest 13 Woodland 1 6 2
Eneabba Nature Reserve 13 Heathland 2 4 1 Boundain Nature Reserve 15 Heathland 7 2 0 Muja Conservation Park 29 Woodland 3 2 2
Tone Perup Nature Reserve 39 Woodland 6 1 1 D'Entrecasteaux National Park
(Northcliffe) 44 Woodland 2 3 1
After collecting leaves for osmotic potential measurements, around midday, five adult
leaves were randomly collected for each species for determination of water content
and leaf mass per area (LMA). Leaves were stored in zip-lock plastic bags in a cooler to
prevent water loss before being weighed in the laboratory approximately 48 hours
later. The area of leaves was measured with a scanner (Epson Perfection V700, NSW,
Australia) after which their fresh and dry weight (after 48 hours in a 60 °C oven) was
recorded. Leaf water content was expressed on a fresh weight basis FW–DW/FW × 100
(Turner, 1981). LMA was calculated by dividing the leaf dry mass by its area (g1m-2).
Soil moisture
Soil moisture was measured in both the burnt and unburnt areas (5 replicates per
area) at each site during both site visits. Soil samples at the burnt and unburnt patches
(5 each patch) were collected from random locations to a depth of 15 centimetres.
This soil was thoroughly mixed and 120 mL of soil from each of the samples was
initially weighed in the laboratory and then dried for 48 hours in 105 °C ovens (Pansu &
Gautheyrou, 2006) and reweighed to determine soil gravimetric water content.
Chapter 5
112
Osmotic potential measurements
Osmotic potential of leaf samples was measured using a vapour pressure osmometer
(PSYPRO CR-7, Wescor Inc, Utah, USA). Sample chambers (C-52, Wescor Inc, Utah,
USA) were cleaned with 50% water: 50% acetic acid and then calibrated using eight
concentrations of standard NaCl salt solutions prior to measurements. Leaf samples
from the -80 °C freezer were allowed to defrost in their vials for one hour before sap
was expressed from leaves (Turner, 1981). Whole leaves were used for smaller leaf
sizes and seedlings whereas the centre of the leaf (excluding midrib) was used for
species, such as Eucalyptus, with larger leaf sizes. Filter paper circles (5 mm in
diameter) were soaked in leaf sap and these samples were transferred into sample
chambers. Based on preliminary measurements, samples were left to equilibrate in
osmometer chambers for 18 minutes before recording leaf osmotic potential.
Data analysis
The statistical package ‘nlme’ in ‘R’ was used for analysis (R Core Team, 2013). A
generalised linear mixed model (GLMM) was constructed to identify significant
correlations among osmotic potential and the main experimental variables (date of
sampling, site, seedlings/mature and all interactions). Species was included as a
random variable nested within site. Site was a categorical variable with seven levels
and therefore deviation coding was used which compared the mean osmotic potential
of each site with the grand mean of all sites (UCLA: Statistical Consulting Group, 2013).
To avoid heterogeneity of variances, osmotic potential values were log transformed.
Initially, maximal models contained all variables within a maximum likelihood
framework which were then reduced using step-wise deletion and the Akaike
Information Criteria (AIC) was used to compare reduced and maximal models.
Diagnostic plots were used to identify heterogeneity in residuals by plotting them
against each explanatory variable. The optimal model had the lowest AIC based on a
chi-squared distribution. From these optimal models, the variable estimates and p-
values were recalculated with a restricted maximum likelihood (REML) which is less
biased than maximum likelihood estimates (Bolker et al., 2009).
Chapter 5
113
Osmotic potential and all climatic variables (AI, rainfall and maximum temperature in
the 7 and 30 days prior to sampling, % soil moisture) were averaged at the community
level and analysed with a linear model. Separate models were created for seedlings
and mature vegetation. Percentage leaf water content and LMA were only included in
the models created for mature vegetation. The optimal model was identified using the
method described above. Gravimetric soil moisture content data were also analysed
with a linear model with site and rainfall data as explanatory variables and percent
moisture content as response variable (Boundain was excluded from this analysis as
moisture could be measured in April only).
The difference in osmotic potential between mature and seedling con-specifics found
at multiple sites were compared using an F-test (to determine differences in variance)
and then a student’s two tailed t-test with a 95% confidence interval. The data from
the two sampling dates were treated separately due to the difference in variance in
osmotic potential between November and April.
RESULTS
Rainfall trends
Most sites experienced rainfall that was slightly lower than the long-term average
before leaf sampling in November 2012, apart from the Eneabba and the Northcliffe
sites which had above-average values (Figure 3). Motague and Boundain, the two sites
furthest inland, both received considerably less rainfall than normal (i.e. less than
70%). In contrast, prior to sampling in April, the four sites with the lowest aridity
indexes (Wandoo, Boundain, Eneabba and Motague) all experienced considerably
higher than average rainfall in the month prior to sampling, whereas the sites with the
highest aridity indexes only had slightly above-average rainfall (Figure 3). As a
consequence, there was no differentiation amongst the experimental sites in terms of
rainfall during the post-summer period. Also, because of these unusual weather
conditions, the sites with the lowest AI received more rainfall in March/ April than in
October/ November, whereas the reverse was true for the sites with the highest AI.
Chapter 5
114
Figure 3 – Total rainfall at the study sites in the 7 and 30 days prior to sampling in (A)
November 2012 and (B) April 2013 compared with the 20 year average during the same time
of the year. Study sites are ordered along the x-axis from the lowest aridity index (AI) to the
highest with the AI provided in brackets Note: Most sites did not have any rainfall 7 days prior
to sampling.
Soil moisture
Soil moisture in the top 15 cm was not significantly related to the amount of rainfall 30
days prior to sampling in either sampling periods (November: R2 = 0.37; d.f. 5; p = 0.2,
April: R2 = 0.13; d.f. 5; p = 0.43). The aridity index of sites was significantly correlated to
soil moisture in April but not November (November: R2 = 0.46; d.f. 5; p = 0.14, April: R2
= 0.76; d.f. 5; p = 0.01). Soil moisture did not differ between pre and post-summer at
the five sites with the lowest AI, even though four of these sites received considerably
Chapter 5
115
more rainfall in April. Tone was the only site with a significant increase in soil moisture
between November and April (d.f. 130; p<0.001; Figure 4), which is likely related to the
substantial rainfall at this site in the week before soil moisture was measured (i.e.
~20mm; Figure 3). Eneabba had the lowest soil water contents which is reflective of
the sandy soils in this coastal heathland habitat. There were no significant differences
in soil moisture between ash and non-ash habitats within the same site (results not
shown).
Figure 4 – Gravimetric soil moisture content ± SE (n = 5) of the upper 15 cm of soil across study
sites during both sampling periods (November 2012 and April 2013). Sites are ordered from
lowest to highest aridity index (values in brackets). Boundain was only sampled in April as a
large storm saturated the soil just after leaf sampling and before soil moisture measurements.
Asterisks indicate significant differences between sampling dates for each site (p <0.05).
Community-averaged leaf osmotic potentials
The mixed model demonstrated significant differences in community-averaged
osmotic potential amongst sites, dates and maturity stage (Figure 5). The wettest
community (Northcliffe) had the highest osmotic potential, while plants in the driest
community (Wandoo) had some of the lowest values. However, there was no clear
relationship between site osmotic potential and aridity index. At most sites,
community-averaged osmotic potentials were significantly less negative in April (d.f.
825; p<0.001). However, this was not the case for vegetation at the Muja site, which
0 2 4 6 8
10 12 14 16 18
Wandoo (12)
Eneabba (13)
Motague (13)
Boundain (15)
Muja (29)
Tone (39)
Northcliffe (44)
Soil
moi
stur
e (%
) Nov-12
Apr-13 *
Chapter 5
116
had similar values during both sampling dates (three way interaction = site, date and
maturity d.f. 825; p = 0.005). Also, the change in osmotic potential between sampling
dates was not related to the aridity index of sites (Figure 5; d.f. 83; p = 0.31). Seedlings
generally had less negative osmotic potentials than adults. It should be noted that at
all sites the species sampled for mature vegetation were not always the same species
as those sampled for seedlings. There were no clear patterns in the osmotic potentials
of seedlings and adults of the same species collected at the same site.
Figure 5 – Community-averaged osmotic potential ± SE of mature plants and seedlings in November 2012 and April 2013. Sites are ordered from the lowest (Wandoo) to highest (Northcliffe) aridity index (AI included in brackets). Asterisks indicate a three way interaction between date, site and maturity stage (d.f. 825; p = 0.005). N = 4, 7, 9, 9, 7, 8 and 6 (mature species) and 4, 8, 9, 8, 5, 5, 6 respectively (seedling species).
Chapter 5
117
Influence of environmental variables and leaf traits on leaf osmotic potentials
For adults, the mixed-model results showed that community-averaged leaf osmotic
potentials were significantly related to date of sampling, leaf and soil water content
(Figure 6: A and C). Leaf osmotic potentials decreased with lower leaf water contents
(i.e. increasing concentration of leaf solutes). Vegetation communities located in sites
with higher soil moisture generally had less negative osmotic potentials. Tone (blue)
was an outlier in all of the regressions for November, with a much lower community-
averaged osmotic potential than would be expected on the basis of leaf water content
and soil moisture. When analysing osmotic potentials of mature individuals for all
species together, species in the low height category generally had a lower osmotic
potential compared with species in medium and high height classes (ave. OP low =
-2.9 MPa, n = 228; medium = -2.5 MPa, n = 179; high = -2.73 MPa n = 81; d.f. 487, p =
0.005).
Chapter 5
118
Figure 6 – Community-averaged leaf osmotic potential (MPa) as related to leaf traits (A:
average% leaf water content, B: LMA) and climatic variables (C: % soil moisture to 15 cm, D:
total rainfall 30 days before sampling, E: aridity index and F: maximum daily temperature 30
days before sampling) for the seven experimental sites in November 2012 (circles) and April
2013 (triangles). Only values for mature individuals are included. Sites are colour coded with
red = Wandoo (lowest AI), orange with black border = Motague, orange with grey border =
Eneabba, yellow = Boundain, green = Muja, Blue = Tone and purple = Northcliffe (highest AI).
P-values were calculated with a chi-squared test during stepwise deletion of variables in the
mixed model. Leaf water content and soil water content were two significant factors in the
mixed model and their R2 were calculated using simple linear regressions.
Chapter 5
119
The community-averaged osmotic potential of seedlings was weakly influenced by
date and rainfall 30 days before sampling (significant interaction, d.f. 79; p = 0.02). The
osmotic potential of seedlings collected post-summer (April 2013) slightly increased
with more preceding rainfall, whereas during the pre-summer sampling (November
2012) there was no such relationship.
Mature and con- specific seedling comparison
For those species where both seedlings and mature plants could be sampled at the
same site, seedlings generally had less negative osmotic potentials (Appendix 6).
Furthermore, mature individuals of species in the low and medium height category
were more likely to have lower osmotic potentials than their con-specific seedlings
(~35% of comparisons, n = 9 species each category) compared to ~9% of species in the
high height category (n = 6 species).
Mature species measured at multiple sites
For the seven species that occurred in at least two sites, it was possible to do a
comparison within species across sites (Figure 7). There was considerable variation
amongst species in leaf osmotic potentials, particularly in the November
measurements. However, populations at the lower AI locations (i.e. drier sites) did not
have consistently lower osmotic potentials and the direction of change was highly
species-specific. There was a tendency for species with taller growth habits (‘H’
symbols in Figure 7) to have lower osmotic potentials at the low AI sites, whereas most
of the smaller shrub species (‘L’ symbols) showed the opposite response. However,
note that the three lowest values obtained were all from populations occurring at the
Tone site. During the April measurements, differences amongst species were smaller
and there was no difference in osmotic potential between con-specifics.
Chapter 5
120
Figure 7 – Leaf osmotic potential ± SE of mature plants (n = 5) for species that occurred at two
of the experimental sites as differentiated by aridity index (= AI in brackets) and measurement
date. Note that Adenanthos cygnorum was located at two sites with the same AI. To provide
an indication of a species growth habit they are categorised by their average height as mature
plants (low: 0–1 metre; mid: 1–2.5 m and high > 2.5 m). Asterisks indicate for each species
whether the osmotic potentials differed significantly for the two sites (p<0.05).
Chapter 5
121
DISCUSSION
This study has provided an overview of the pre-and post-summer dynamics of leaf
osmotic potential across an aridity gradient. There was significant variation in
community-averaged leaf osmotic potential amongst sites, dates and plant maturity
stages, and osmotic potential values were highly species-specific. Thus, they may not
be valuable as stand-alone measures to assess a species’ drought tolerance and
potential response to a drying climate.
Community variation along an aridity gradient
Contrary to the meta-analysis by Bartlett et al. (2012), sites in drier climates (lower AI)
did not show consistently lower community-averaged leaf osmotic potentials. There
are several reasons that could explain this apparent discrepancy. Firstly, AI was not a
good estimate of short-term water availability and may not be a good estimate of
longer term water availability. Ideally, sites would have been located in similar
landscape positions with similar vegetation communities and soil types, as all these
aspects would influence local water availability. This was not the case as only a limited
number of post-fire sites could be located, and variation in soil type (e.g. sandy soils at
Eneabba versus clay soils at Wandoo) and vegetation community (woodland versus
heathland) were apparent. Secondly, the weather conditions on the days of sampling
may have introduced variation in the osmotic potential readings among sites. For
example, warmer days with higher vapour pressure deficits could have led to faster
leaf water loss resulting in more negative osmotic potentials. Thirdly, the timeframe
over which the leaves were sampled at each site may have increased variability in
osmotic potential readings. Leaf samples of all species were taken within a four hour
timeslot, and although samples were stratified according to species, leaves that were
sampled later could have lost more water resulting in a lower osmotic potential and an
increase in within species variability. Thus, future measurements could standardise
osmotic potential measurements by saturating leaf samples to full turgor before
osmotic potentials are determined, or by determining leaf relative water content on a
Chapter 5
122
separate set of leaves. The latter would enable calculation of osmotic potential at full
turgor. This would also provide insight into whether plants are using passive or active
osmoregulation in response to soil drying (Hsiao et al., 1976).
Despite the above shortcomings, differences in osmotic potential did demonstrate
some adaptive responses and more short-term plastic responses to rainfall. Consistent
differences in osmotic potential were evident at the extreme ends of the aridity
gradient (Northcliffe and Wandoo) in all correlations with climate and leaf traits. Both
of these sites were in a woodland habitat and had a similar range of species with
respect to species height categories. Therefore, the large difference in community-
averaged osmotic potentials between Northcliffe and Wandoo suggest an adaptive
response to local water availability. Similarly, Grieve and Hellmuth (1970) found a
gradient in osmotic potential in temperate to arid vegetation across a rainfall gradient
from Dwellingup (annual rainfall ~1316 mm) to Cue (~200 mm), Western Australia. In
contrast, Tone was the site with the second highest aridity index and the community-
averaged osmotic potentials exhibited short-term plasticity to local water availability
between sampling dates. In November, the average osmotic potential at this site was
-4 MPa which was an outlier across all regressions of climate and leaf traits. This
suggests that mature vegetation at Tone were stressed due to the below average
rainfall before sampling. The average osmotic potential at this site showed the
greatest increase (35%) compared with all other sites after substantial rainfall in April.
While adaptive responses can be important to survive in a drying climate, it has been
demonstrated that in some cases the rate of species evolution will not keep up with
rapid environmental change (Jump & Peñuelas, 2005). Therefore, short term plasticity,
such as osmotic changes measured at the Tone site, might be an important mechanism
for plants to survive longer drought periods.
Seasonal variation
The osmotic potential of mature vegetation varied significantly between sampling
dates, with most sites having lower osmotic potentials in pre-summer (November
Chapter 5
123
2012) compared to post-summer (April 2012). This result is contrary to expectation
and is likely due to the unusual weather conditions with considerably higher than
average rainfall before the April sampling, especially at the drier sites. The higher
rainfall did not lead to measurable differences in top soil moisture, apart from at the
Tone site which was the only site with significant rainfall in the week before sampling
(Figure 4). It is likely that the upper soil layer had dried at most sites, and that the less
negative osmotic potentials in April were the result of tissue rehydration in response
to above average late summer rainfall and possibly higher soil water contents deeper
down the soil profile.
Correlations with other traits
Community-averaged osmotic potentials of mature vegetation were significantly
correlated with average leaf water content. As expected, vegetation communities with
more negative osmotic potentials tended to have lower leaf water contents. As
mentioned earlier, concurrent measurements of relative water content and osmotic
potential from leaves collected at the same time would be essential to determine if
these patterns (1) were related to active regulation of leaf solutes to tolerate drought,
and (2) whether the values obtained reflect true differences amongst the communities
rather than differences in leaf water content. In contrast, LMA did not have a
significant relationship with community-averaged osmotic potential. This lack of
correlation concurs with the meta-analysis of biomes by Wright et al. (2004) and
Bartlett et al. (2012). Leaf mass per area is known to vary within a species across
environmental gradients and also on a temporal scale (Niinemets, 2001). Moreover, in
another global review, Poorter et al. (2009) found LMA varied most strongly with light
availability followed by temperature and submergence, while water availability had a
much smaller effect. Therefore, differences in LMA are probably driven by other
environmental variables, not just those related to soil water availability.
The height of mature individuals of a species was weakly related to their osmotic
potential with shorter species generally having lower osmotic potentials. Plant height
Chapter 5
124
typically scales with belowground biomass (Enquist & Niklas, 2002); therefore the
lower osmotic potentials of shorter species may be a reflection of a shallow root
system and consequently a reduced access to water. Indeed, Mitchell et al. (2008)
demonstrated that shallow-rooted species in the woodlands of Western Australia
changed their osmotic potential to a greater extent than deeper rooted ones in
response to low water availability to acquire more water from the dry soils. Dodd et al.
(1984) and Groom (2004) also demonstrated that rooting depth was correlated with
drought tolerance or avoidance responses on the coastal sandplains of Western
Australia. These results suggest that measuring root depth or shoot height does assist
with understanding drought tolerance at a species or community level in southwest
Australia.
Seedling responses
Most environmental variables did not correlate with osmotic potential apart from a
significant interaction between date and rainfall 30 days prior to sampling. In April
2013, the osmotic potential of seedlings was more closely correlated with rainfall
compared with November 2012 sampling. The lack of correlation is an unexpected
result, as seedlings usually respond directly to local rainfall and soil moisture compared
with mature vegetation that may have access to ground water (Niinemets, 2010).
These results could be attributed to the high variation in osmotic responses in
seedlings. In addition to the sources of variation described above, comparisons
between seedlings were further complicated due to the different stages of maturity
i.e. some seedlings would have had more mature leaves that others and some
seedlings would have had deeper roots. Plasticity in rooting depth has been
significantly correlated to seedlings mortality due to drought in woody shrub species
north-east Spain (Padilla et al., 2007) and 42 shrub species in Britain (Reader et al.,
1993). Therefore, information about root depth of seedlings could complement
osmotic potential measurements to further understand the thresholds of drought
severity associated with failed and successful seedling establishment.
Chapter 5
125
Comparisons at the community level revealed that seedlings often had significantly
less negative osmotic potentials compared with mature vegetation. This pattern was
also evident between mature and seedlings of the same species and could be
attributed to two mechanisms. Firstly, while mature plants of many species are able to
increase their water intake by increasing cellular solutes (i.e. osmotic adjustment), this
mechanism is not fully developed in growing tissue of seedlings. As seedlings do not
have fully developed cell walls, drawing in more water through osmotic adjustment
would further increase cell expansion thereby increasing the seedlings evaporative
demand which would be counterproductive. Therefore, seedlings have to keep their
osmotic potentials relatively high (Davis & Mooney, 1986). A second explanation for
the seedlings higher osmotic potentials may be related to the potentially higher water
availability in the burnt versus the unburnt habitats. For example, Mappin et al. (2003)
found that seedlings in an arid shrubland community in Western Australia that had
established in recently (5 yr) burnt sites were less drought stressed than long unburnt
vegetation nearby. Seedlings at the burnt site had higher stomatal conductance and
less negative water potentials, which was attributed to high soil moisture at the burnt
site compared to the unburnt sites. A combination of these processes could explain
differences in leaf osmotic potential between mature and seedling in the present
study. While soil moisture to 15 cm did not differ between burnt and unburnt areas at
the same site (results not shown), seedlings had probably developed roots deeper than
15 cm before the onset of summer. For example, Enright and Lamont (1992) showed
that roots of Banksia seedlings can reach up to 2 m in deep sands within their first
year. Thus, both differences in leaf physiology between mature plants and seedlings as
well as differences in water availability between burnt and unburnt habitats at the
same locations, could potentially explain the generally higher osmotic potentials of
seedlings compared with mature plants.
Implications for the prediction of community responses to a drying climate
Based on the lack of correlations between leaf osmotic potential and environmental
variables, and the large variation in osmotic potential among species within sites, it is
Chapter 5
126
clear that more information is required to determine differences in drought tolerance
amongst common or threatened species and communities. Evidently, measuring just
one physiological adaptation to drought, osmotic potential, is not enough to
determine a species vulnerability to a drying climate. For example, observing a very
negative leaf osmotic potential could indicate that a plant is severely drought-stressed
but could equally well be indicative of a plant that is extremely well-adapted by still
being able to extract water from a very dry soil. Other concurrent measures, such as
leaf transpiration rates or stomatal conductance, would allow discrimination between
these possibilities. Also, in highly heterogeneous landscapes, such as in southwest
Australia, other site variables that determine plant water availability such as
topography in the landscape, the soil profile and texture and species traits such as
rooting depth would further assist in making more robust predictions about local
species drought tolerances. Similar to this study, other studies conducted in
Mediterranean climates have found a similarly high variation in physiological
responses of co-existing mature vegetation in the field in terms of water potential,
photochemical efficiency and leaf proline content (Zunzunegui et al., 2011), as well as
hydraulic conductivity, stomatal conductance, photosynthetic measures and
morphological traits (stem area and leaf area) (Quero et al., 2011). Burgess (2006) also
emphasised the need to understand species water relations for revegetation in the
wheatbelt region of SW Australia but highlighted the difficulties doing this in highly
diverse landscapes. Understanding the water relations in a limited number of strategic
vegetation communities (i.e. either threatened or dominant communities harbouring
high biodiversity) may be a more achievable goal when predicting climate-related
changes in highly diverse environments.
Chapter 6
127
Chapter 6
General Discussion
My thesis has explored the factors that drive seedling establishment in natural and
translocated populations of Acacia and Banksia species at three spatial scales. The
studies detailed in the preceding chapters were directed towards informing the
establishment of future translocations in southwest Australia and other
Mediterranean-climate regions. This concluding chapter discusses the main outcomes
of my thesis and potential future research directions within the field of translocation
science, and with particular emphasis on translocations in a drying climate.
THESIS AIMS
Is it possible to manipulate microhabitat “safe site” variables and plant seedlings with
particular characteristics that together improve their growth and survival in a
Mediterranean-climate?
Chapter 2 and 3 discuss the effects of manipulations at the seedling and microhabitat
level within translocations of Acacia awestoniana and Banksia ionthocarpa subsp.
ionthocarpa. Translocated populations were not only established to reduce the risk of
extinction of both species but also to learn more about the importance of planting in
and manipulating different microhabitats, and to ameliorate environmental factors
that might affect seedling survival in Mediterranean-climate regions. Another key
objective was to further understand the ecology of the species, as well as their
response to environmental factors including the plasticity of their response that might
in turn effect growth and survival.
Both translocations had a relatively high two-year survival (81% and 75% respectively)
which greatly increased the individual number of plants that currently exist in the wild.
The translocation of A. awestoniana increased the population size by ~170%, while
there was a ~62% increase in individuals for B. ionthocarpa. Furthermore, some Acacia
awestoniana seedlings had already reached reproductive maturity in the second spring
Chapter 6
128
since planting. However, high survival at early establishment stages after translocation
does not guarantee long term persistence. For example, Drayton & Primark (2000)
established translocations of eight perennial species between 1994 and 1995 in
eastern Massachusetts, USA. Two years later, seven of these translocated species were
established and all had produced fruit. However, 15 years after planting only two
species were present and producing fruits (Drayton & Primack, 2012).
By establishing a multi-level experiment (i.e. manipulations at the seedling and plot
level) within both translocations, I was able to disentangle the relative importance of
different factors in the translocation process. Pre-translocation substrate and planting
season were dominant drivers influencing survival. The effect of planting in different
seasons appeared more related to seedling size at planting than through a direct effect
of season itself. Seedlings that were larger at planting had a consistent survival
advantage for both species. I demonstrated that root length at planting was
significantly correlated with survival in B. ionthocarpa, where seedlings with 15–20 cm
long roots had an 80% greater chance of survival compared with seedlings with shorter
roots. Secondly, both translocations confirm the advantage of planting seedlings with
an intact potting mix surrounding the roots. This was generally the fastest method of
planting large numbers of plants and presumably reduced the chance of root damage.
It is likely that the root systems of these seedling experienced less ‘transplant shock’
and profited from initially being more buffered from the external environment.
Experimental manipulation at the seedling level has not been commonly incorporated
into translocations to date, but my research demonstrates that it would be worthwhile
as seedling characteristics can strongly influence early establishment success.
Manipulating the characteristics of safe sites within both translocations, i.e. water
availability and the potential level of competition with surrounding vegetation,
revealed that the relative safeness of a microhabitat varies with seedling traits.
Generally, the effects of less favourable seedling characteristics were exacerbated in
sub-optimal microhabitats. For example, in B. ionthocarpa, there was significantly
lower survival of bare rooted seedlings planted in the tall heath plots watered on a
monthly basis. The combination of potential root damage, low root-soil contact and a
Chapter 6
129
thick litter layer led to highly vulnerable seedlings. In addition, the younger and smaller
A. awestoniana seedlings had a significantly lower growth rate when planted under E.
wandoo with no watering. My findings suggest that larger seedlings with longer root
systems will have higher establishment success growing in a range of optimal and sub-
optimal microhabitats. Thus, although more costly, growing seedlings to larger sizes
before transplanting them, and possibly in larger pots to allow for more root growth,
may be most beneficial.
Do plant physiological responses and microhabitat measurements help to explain
seedling survival and growth?
Monitoring of survival and growth of A. awestoniana (Chapter 2) and B. ionthocarpa
(Chapter 3) was complemented with measures of environmental variables and
seedling health over the two year period. I used two different methods to measure
relative soil moisture and in general found that the length of time between watering
and moisture measurements was too long to detect differences among watering
treatments. This suggests that the supplemented water had dissipated from the upper
soil layers within a week. However, at the A. awestoniana translocation site I was able
to couple neutron moisture measurements on a monthly basis with measures of
seedling transpiration using the infrared camera. This enabled me to suggest a possible
sequence of events in response to water supplementation. Twenty four hours after
watering, the leaves of seedlings were significantly cooler than those in no watering
plots. This demonstrates that supplemented water in the upper layers was actively
being used by seedlings. A week after watering, plots watered on a monthly and
weekly basis had significantly higher relative soil moisture at intermediate depth (60 –
100 cm) compared to non-watered plots. It is likely that the supplemented water had
reached the intermediate layer by this time. However, the roots of seedlings may not
have reached this depth yet, or only had few roots established in this horizon. The
combined information on seedling growth, soil moisture and transpiration suggests
that growth of A. awestoniana seedlings over summer was largely reliant on the
weekly increases in moisture which was only short-lived in the upper soil layers.
Chapter 6
130
Differences in soil moisture measurements from both translocations could also be
linked to the effect of surrounding vegetation. Relative soil moisture measured in the
top 15 cm at the B. ionthocarpa site provided evidence that bare plots retained higher
soil moisture in the upper layers compared with low and tall heath plots. However,
other environmental factors such as shallow soil depth and potentially exposed
conditions made the bare microhabitat unfavourable for seedling survival. At the A.
awestoniana site, the presence of large E. wandoo trees could explain the increased
relative soil moisture in the deeper layers (i.e., > than 100 cm). Possibly E. wandoo
roots were transporting irrigation water from upper depths to deeper soil layers along
a water potential gradient, which would have negatively impacted on seedlings. Similar
measures of soil moisture were used by Roncal et al. (2012) to assess optimal
microhabitats after transplanting the endangered Amorpha herbacea var. crenulata
outside its historic range. Historically, this species was found in a habitat between the
pine rockland and grassland, and yet transplants had the highest survival in pine
rockland which was attributed to lower vegetation cover and higher soil moisture.
Subsequently, a subset of seedlings was moved from the historical habitat to the pine
rockland which significantly increased overall survival. Measuring soil moisture during
seedling establishment can assist with understanding the effects of various treatments
as well detecting possible competitive effects of surrounding vegetation.
Other environmental variables (i.e. percentage regrowth of surrounding vegetation
and soil surface temperature) and seedling health and physiology (i.e. transpiration,
percentage herbivory and yellowing of leaves) complemented monitoring of survival
and growth in both translocations. These measures helped to understand seedling
responses to microhabitat in more detail compared with simple measures of survival
and growth, which further helped to define safe sites for these species. The
significance of multi-faceted approach was demonstrated at the B. ionthocarpa
translocation, where measures of regrowth in low heath plots revealed that seedlings
with regrowth present in their immediate vicinity had a significant survival advantage
after two years. Longer term monitoring will be required to determine if this regrowth
has a competitive effect on seedlings after watering ceases. In addition, quantifying
herbivory at the A. awestoniana site indicated that larger seedlings were a significantly
Chapter 6
131
greater target for herbivores when located under E. wandoo compared to smaller
seedlings or being located in the open. By incorporating all measures, I concluded that
optimal safe-sites in the early establishment stages were in the low heath
microhabitats for B. ionthocarpa which often had regrowth and in an open sunny
microhabitat for A. awestoniana. Thus, using a comprehensive experimental design
and a range of environmental measures within a translocation can provide detailed
recommendations to land managers for the future.
Can we use data on natural seedling establishment and physiological responses to
guide conservation of threatened species in a drying climate?
I surveyed common species in natural settings in Chapters 4 and 5 to further
investigate safe site preferences for seedling survival and growth in a post-fire
environment and changes in osmotic potential across a climatic gradient. Investigating
natural recruitment could provide a better understanding about how to ameliorate
drought for future translocations projects.
Some of the findings from the seedling establishment survey (Chapter 4) were already
being implemented within the translocation program in Western Australia. Firstly,
herbivory was the main biotic factor constraining seedling survival and fencing has
been used within translocations to overcome this problem. The survey also provided
some insights into other methods that could be used to reduce the risk of herbivory
such as planting seedlings under dead branches or logs. Microtopography was another
factor influencing survival of tagged seedlings, with survival significantly lower on
convex surfaces. This could be due to a combination of a lack of rainfall capture and
increased herbivory in these locations. Depressions were created before planting B.
ionthocarpa seedlings and have the potential to improve early survival in other
translocations. Increased herbivory is a potential threat to planted seedlings in a drying
climate as surrounding mature vegetation dies and higher densities of herbivores could
potentially use post-fire sites for food. Therefore implementing a range of herbivore
Chapter 6
132
controls will probably become a priority for translocations in Mediterranean-climate
regions.
The post-fire surveys (Chapter 4) and osmotic potential measurements (Chapter 5)
revealed some features of a post-fire environment that could be manipulated within a
translocation site. For example, microhabitats with ash were found to improve growth
of Banksia species at the Stirling Range site (Chapter 4). Creating ash beds was
successful in the restoration of Eucalyptus gomphocephala in southwest Australia and
therefore has the potential to assist with growth and survival in translocations of
threatened species (Ruthrof et al., 2010). Furthermore, burning patches before
planting reduces surrounding competition (plus thick litter layers if present) which may
improve water availability for seedlings. Increased water availability in post-fire sites
could partly explain the less negative osmotic potentials of seedlings compared with
con-specific mature vegetation in nearby unburnt habitats (Chapter 5). Burning has
been used as a treatment to promote seedling establishment in other projects in
Mediterranean regions. Holmgren et al. (2000) found consistently higher survival and
growth eight years after planting woody seedlings in burnt habitats compared to
cleared and intact habitats in the Chilean matorral. Buist (2003) compared the survival
of rare and common Acacia species when planted into burnt and unburnt habitats in
the wheatbelt region of Western Australia. All species had consistently higher survival
in the burnt habitats which was predominately attributed to greater access to light and
water. Therefore, incorporating some post-fire characteristics into translocations has
the potential to improve seedling establishment.
When incorporating post-fire characteristics into a translocation site, it is important to
consider that large-scale clearing before planting may also hinder seedling
establishment. Results from the B. ionthocarpa translocation (Chapter 3) and post-fire
surveys (Chapter 4) suggest that mature vegetation or regrowth can ameliorate less
favourable conditions in certain circumstances. This result could be due to improving
the local water availability/ temperature as for Banksia ionthocarpa seedlings in low
heath plots or providing refugia from herbivores in a post-fire environment. Gomez-
Aparicio et al. (2004) found positive effects of existing vegetation after establishing a
Chapter 6
133
large-scale restoration experiment in another Mediterranean-climate region
(southeast Spain). After planting eleven species of seedlings either in the open or
under nurse vegetation in a range of habitats, 75% of experimental cases showed
significantly higher survival under nurse shrubs. Huggins et al. (2010) and García &
Ramón Obeso (2003) also used nurse plants to reduce herbivory and improve water
availability for two threatened species Astragalus jaegerianus (granite outcrop species
in California) and Taxus baccata (North-west Spain). Nurse plants can therefore help
protect seedlings from projected increasing drought and temperatures (Padilla &
Pugnaire, 2006). Nurse plants have the potential to be increasingly important when
planting seedlings in southwest Australia and other Mediterranean-climate regions.
More data are needed to understand the role of nurse plants in southwest Australia,
where their use in translocations/ restorations have received less attention than in
other Mediterranean-climate regions.
RECOMMENDATIONS
My thesis has demonstrated that there are a range of key factors that can lead to
higher survival and growth within translocations and I provide the following
recommendations for future translocations in a drying climate.
Larger seedlings with longer root systems will have higher success growing in a
range of optimal and sub-optimal microhabitats and future translocations should
consider methods to best produce large seedlings before planting.
Combining seedling survival and growth monitoring with other measures of the
surrounding environment and seedling morphology/ physiology can assist with
understanding species ecology and plasticity to their microhabitat to guide
locations for planting in future translocations.
Infrared thermography can be used as a measure of seedling stress to drought. It
could be used at critical time-points within the translocation to assess seedling
stress and to support management decisions. For example, infrared thermography
might be used as a guide to determine when water should cease after a summer
Chapter 6
134
drought. Infrared thermography might be particularly useful for assessing the
stress of seedlings translocated into areas outside their historical distribution.
Herbivory has a negative effect on planted seedlings, and therefore controlling
herbivory is a priority for translocations in southwest Australia.
Nurse plants have a potential role to assist with establishing translocated seedlings
in southwest Australia as other studies have shown in other Mediterranean-climate
regions.
Some aspects of post-fire environments, such as ash beds and burning to increase
soil moisture, could be implemented to improve seedling establishment within
translocations in southwest Australia.
FUTURE RESEARCH DIRECTIONS
Much further research is required to continue improving our understanding of the
main factors that limit seedling establishment and survival to reproductive maturity in
Mediterranean-climate regions. Both translocations demonstrated that seedling size
(Chapter 2 and 3) and root size (Chapter 3) were main determinants of seedling
survival. In order to maximise seedling establishment, further research should focus on
optimal plant and root sizes for translocation, which is likely to be partly species and
habitat dependent, and optimal pre-translocation growth conditions. The latter may
involve pre-conditioning seedlings for faster acclimation to warm and dry
environments. In both translocation experiments, information on the development of
the seedling root system as dependent on watering regimes and microhabitats would
have been highly beneficial. However, in most situations harvesting of rare species
would not be acceptable. Therefore, future research should incorporate using
surrogate species in similar experimental settings. In particularly, closely related
species occurring in a similar habitat as a rare target species, or local species in the
same functional group, would be good candidates.
Results from this thesis suggest that seedlings can benefit from a lack of surrounding
competition (Chapter 2) and possibly increased water availability after burning
Chapter 6
135
(Chapter 4), while other trends suggest that mature vegetation can play a role in
ameliorating abiotic conditions (Chapter 3). Manipulating the effects of surrounding
vegetation in an experimental manner will help to disentangle the competitive vs.
facilitative effects. Measuring soil moisture on a daily basis at strategic time points
with a neutron moisture probe can provide detailed data about the movement of
supplemented water through the soil profile. This could be coupled with root
excavations to gain insight into the morphological responses of seedlings with and
without surrounding vegetation.
The combination of long lifespan and widespread fragmentation means that many
threatened plant species may not be able to keep up with rapid changes in climate.
Consequently, efforts to increase population sizes using reintroduction/ augmentation
methods could be hindered if species cannot tolerate climatic changes in their current
range. Therefore, some researchers have recommended moving species outside their
native range through a process known as assisted migration (McIntyre, 2011; Thomas,
2011). To date, assisted migration has been rarely reported in the conservation of
flora within Australia. McIntyre (2011) assessed the viability of using assisted migration
to conserve threatened grassy woodland communities in southeast Australia. This
study concluded that introducing threatened grassland species could be incorporated
into revegetation practices that were already well-established the region to improve
overall vegetation connectivity. However, Regan et al. (2012) demonstrated that
assisted migration is not a useful management option if other prevailing threats such
as inappropriate fire regimes are not addressed simultaneously. Ultimately, cases need
to be assessed individually and assisted migration, like translocations, could perhaps
be carried out on an experimental basis with intensive long term monitoring. Burbidge
et al. (2011) and Harris et al. (2013) provide detailed evaluations about considering
assisted migration in Australia.
Comparisons of osmotic potential in Chapter 5 provided some indication that species
and plant communities are highly dynamic in their responses to changes in climate. It is
recommended that other physiological responses such as stomatal conductance, leaf
Chapter 6
136
water potential and vulnerability to xylem cavitation are measured during different
seasons across communities in southwest Australia. This will provide a comprehensive
understanding of water relations along a climatic gradient that could be incorporated
into predictions for future distribution. For example, McDowell et al. (2013)
incorporated physiological data and models at different spatial scales to predict
mortality in piñon pine–juniper woodland. The predictive power of these complex
models will continue to become more accurate and will assist with conservation efforts
across the globe.
CONCLUSIONS
I have used surveys of natural systems as well as manipulated experiments to gain a
comprehensive understanding about seedling establishment in Acacia and Banksia
species in southwest Australia. Both of these methods had benefits and limitations. For
example, surveys such as those in Chapter 4 and 5, allowed the measurement of many
environmental variables across at a range of sites to provide an overview of the main
patterns of seedling establishment and physiological responses. However, the results
were completely dependent on climatic and environmental variation in the survey
period and results were only correlative. By implementing manipulations in the field
(i.e. Chapter 2 and 3), I was able to control the species, site and some environmental
conditions. Experimental manipulations can provide stronger conclusions about some
of the effects of the imposed treatments. However, sometimes interactions between
experimental treatments can be complex, such as the interactions between seedling
and microhabitat levels on A. awestoniana and B. ionthocarpa growth and survival.
Thus, the relative importance of various treatments can be difficult to disentangle.
Taken together, all chapters highlighted that the “one size fits all” approach does not
always work in translocations.
Measuring translocated seedling responses after planting (Chapter 2 and 3), natural
seedlings responses after fire (Chapter 4) and measuring the osmotic potential of
species (adults and seedlings) occurring in a range of different vegetation communities
Chapter 6
137
(Chapter 5) all demonstrated that individuals within a species and among species have
a wide range of physiological and/or morphological responses to their external
environment. This variability in responses, or phenotypic plasticity, is essential for
survival in response to rapid environmental change such as a drying climate (Nicotra et
al., 2010). However, this plasticity makes it difficult to assign a set of safe site
characteristics to all species being considered for translocation in southwest Australia
or other Mediterranean-climate regions. Safe sites can vary with seedling size at
planting, particularly when in competition with a canopy species (Chapter 2) and biotic
pressures, such as herbivory, can dominate at one site but not evident at another
(Chapter 4). My thesis has implemented a range of methods to further understand safe
site variables and methods to grow robust seedlings that have a better chance of
surviving a drying climate. The recommendations arising from this research have the
potential to improve early establishment in southwest Australia and other
Mediterranean-climate regions. Thus, my thesis has made an important contribution
to the current worldwide effort to improve the success of threatened plant
translocations.
References
138
References
Abrams, M. D., Kubiske, M. E., & Mostoller, S. A. (1994). Relating wet and dry year ecophysiology to leaf structure in contrasting temperate tree species. Ecology, 75(1), 123-133.
Adams, M. A., Simon, J., & Pfautsch, S. (2010). Woody legumes: A (re)view from the south. Tree Physiology, 30(9), 1072-1082.
Aguilera, M. O., & Lauenroth, W. K. (1995). Influence of gap disturbances and type of microsites on seedling establishment in Bouteloua gracilis. Journal of Ecology, 83(1), 87-97.
Albrecht, M. A., & McCue, K. A. (2010). Changes in demographic processes over long time scales reveal the challenge of restoring an endangered plant. Restoration Ecology, 18, 235-243.
Alexander, L. V., & Arblaster, J. M. (2009). Assessing trends in observed and modelled climate extremes over Australia in relation to future projections. International Journal of Climatology, 29(3), 417-435.
Allen, C. D., Macalady, A. K., Chenchouni, H., Bachelet, D., McDowell, N., Vennetier, M., Kitzberger, T., Rigling, A., Breshears, D. D., Hogg, E. H., Gonzalez, P., Fensham, R., Zhang, Z., Castro, J., Demidova, N., Lim, J.-H., Allard, G., Running, S. W., Semerci, A., & Cobb, N. (2010). A global overview of drought and heat-induced tree mortality reveals emerging climate change risks for forests. Forest Ecology and Management, 259(4), 660-684.
Allen, W. H. (1994). Reintroduction of endangered plants. Bioscience, 44(2), 65-68. Allred, B. W., Fuhlendorf, S. D., Engle, D. M., & Elmore, R. D. (2011). Ungulate
preference for burned patches reveals strength of fire–grazing interaction. Ecology and Evolution, 1(2), 132-144.
Anderegg, W. R. L., Anderegg, L. D. L., Sherman, C., & Karp, D. S. (2012). Effects of widespread drought-induced aspen mortality on understory plants. Conservation Biology, 26(6), 1082-1090.
Arrieta, S., & Suárez, F. (2006). Marginal holly (Ilex aquifolium L.) populations in Mediterranean central spain are constrained by a low-seedling recruitment. Flora - Morphology, Distribution, Functional Ecology of Plants, 201(2), 152-160.
Babu, R. C., Pathan, M. S., Blum, A., & Nguyen, H. T. (1999). Comparison of measurement methods of osmotic adjustment in rice cultivars. Crop Science., 39(1), 150-158.
Bannister, J., Coopman, R., Donoso, P., & Bauhus, J. (2013). The importance of microtopography and nurse canopy for successful restoration planting of the slow-growing conifer Pilgerodendron uviferum. Forests, 4(1), 85-103.
Barberis, I. M., Batista, W. B., Pire, E. F., Lewis, J. P., & León, R. J. C. (2002). Woody population distribution and environmental heterogeneity in a chaco forest, Argentina. Journal of Vegetation Science, 13(5), 607-614.
Barrett, S., Dillon, R., Cochrane, A., & Monks, L. (2011). Conservation introduction of a threatened narrow range endemic species, Banksia montana, from the Stirling Range National Park, southern Western Australia (Global Re-introduction
References
139
Perspectives: 2011. More case studies from around the globe.). Gland, Switzerland: IUCN/SSC
Barrow, N. (1977). Phosphorus uptake and utilization by tree seedlings. Australian Journal of Botany, 25(6), 571-584.
Bartlett, M. K., Scoffoni, C., & Sack, L. (2012). The determinants of leaf turgor loss point and prediction of drought tolerance of species and biomes: A global meta-analysis. Ecology Letters, 15(5), 393-405.
Battaglia, M., & Reid, J. (1993). The effect of microsite variation on seed-germination and seedling survival of Eucalyptus delegatensis. Australian Journal of Botany, 41(2), 169-181.
Bazzaz, F. A. (1996). Plants in changing environments: Linking physiological, population and community ecology. Great Britain: Cambridge Unviersity Press.
Bell, D. (2001). Ecological response syndromes in the flora of southwestern Western Australia: Fire resprouters versus reseeders. The Botanical Review, 67(4), 417-440.
Bolker, B. M., Brooks, M. E., Clark, C. J., Geange, S. W., Poulsen, J. R., Stevens, M. H. H., & White, J.-S. S. (2009). Generalized linear mixed models: A practical guide for ecology and evolution. Trends in Ecology & Evolution, 24(3), 127-135.
Bradstock, R. (1991). The role of fire in establishment of seedlings of serotinous species from the sydney region. Australian Journal of Botany, 39(4), 347-356.
Brodribb, T., & Hill, R. (1993). A physiological comparison of leaves and phyllodes in Acacia melanoxylon. Australian Journal of Botany, 41(3), 293-305.
Buhk, C., Meyn, A., & Jentsch, A. (2007). The challenge of plant regeneration after fire in the Mediterranean basin: Scientific gaps in our knowledge on plant strategies and evolution of traits. Plant Ecology, 192(1), 1-19.
Buisson, E., Anderson, S., Holl, K. D., Corcket, E., Hayes, G. F., Peeters, A., & Dutoit, T. (2008). Reintroduction of Nassella pulchra to california coastal grasslands: Effects of topsoil removal, plant neighbour removal and grazing. Applied Vegetation Science, 11(2), 195-204.
Buist, M. L. (2003). Comparative ecology and conservation biology of two critically endangered acacias (Acacia lobulata and A. sciophanes) and two common, widespread relatives (Acacia verricula and A. anfractuosa) from the south-west of Western Australia. Unpublished Doctor of Philosophy, University of Western Australia, Perth.
Buist, M. L., Coates, D. J., & Yates, C. J. (2002). Rarity and threat in relation to the conservation of Acacia in Western Australia. Conservation Science of Western Australia, 4(3), 36-53.
Burbidge, A., Byrne, M., Coates, D., Garnett, S. T., Harris, S., Hayward, M. W., Martin, T. G., McDonald-Madden, E., Mitchell, N. J., Nally, S., & Setterfield, S. A. (2011). Is Australia ready for assisted colonization? Policy changes required to facilitate translocations under climate change. Pacific Conservation Biology, 17(3).
Bureau of Meterology. (2013). North Stirlings- monthly rainfall, Climate Data Online: Commonwealth of Australia.
Burgess, S. S. O. (2006). Measuring transpiration responses to summer precipitation in a Mediterranean climate: A simple screening tool for identifying plant water-use strategies. Physiologia Plantarum, 127(3), 404-412.
References
140
Burmeier, S., & Jensen, K. (2009). Experimental ecology and habitat specificity of the endangered plant Apium repens (jacq.) lag. at the northern edge of its range. Plant Ecology & Diversity, 2(1), 65 - 75.
Callaway, R. M., & Walker, L. R. (1997). Competition and facilitation: A synthetic approach to interactions in plant communities. Ecology, 78(7), 1958-1965.
Carey, C. (2005). How physiological methods and concepts can be useful in conservation biology. Integrative and Comparative Biology, 45(1), 4-11.
Carrington, M. E., & Keeley, J. E. (1999). Comparison of post-fire seedling establishment between scrub communities in Mediterranean and non-Mediterranean climate ecosystems. Journal of Ecology, 87(6), 1025-1036.
Castro, J., Allen, C. D., Molina-Morales, M., Marañón-Jiménez, S., Sánchez-Miranda, Á., & Zamora, R. (2011). Salvage logging versus the use of burnt wood as a nurse object to promote post-fire tree seedling establishment. Restoration Ecology, 19(4), 537-544.
Castro, J., Zamora, R., Hódar, J. A., & Gómez, J. M. (2005). Alleviation of summer drought boosts establishment success of Pinus sylvestris in a Mediterranean mountain: An experimental approach. Plant Ecology, 181(2), 191-202.
Caughley, G., Brown, B., & Noble, J. (1985). Movement of kangaroos after a fire in mallee woodland. Australian Wildlife Research, 12, 349 - 353.
Center for Plant Conservation. (2013). International Reintroduction Registry. Retrieved 11th November 2013, from http://www.centerforplantconservation.org/reintroduction/MN_ReintroductionEntrance.asp.
Chambers, D., & Attiwill, P. (1994). The ash-bed effect in Eucalyptus regnans forest: Chemical, physical and microbiological changes in soil after heating or partial sterilisation. Australian Journal of Botany, 42(6), 739-749.
Chanasyk, D. S., & Naeth, M. A. (1996). Field measurement of soil moisture using neutron probes. Canadian Journal of Soil Science, 76, 317-323.
Chapman, A. R., & Gioia, P. (2013). Census of the vascular plants of Western Australia. Kensington, Western Australia: Department of Parks and Wildlife.
Charles, S., Silberstein R, Teng J, Fu G, Hodgson G, Gabrovsek C, Crute J, Chiew FHS, Smith IN, Kirono DGC, Bathols JM, Li LT, Yang A, Donohue RJ, Marvanek SP, McVicar TR, Van Niel TG and Cai W. (2010). Climate analyses for south-west Western Australia. A report to the Australian government from the CSIRO south-west Western Australia sustainable yields project. Australia: CSIRO.
Clemens, J., & Jones, P. G. (1978). Modification of drought resistance by water stress conditioning in Acacia and Eucalyptus. Journal of Experimental Botany, 29(4), 895-904.
Clemente, A. S., Rego, F. C., & Correia, O. A. (2005). Growth, water relations and photosynthesis of seedlings and resprouts after fire. Acta Oecologica, 27(3), 233-243.
Close, D. C., & Davidson, N. J. (2003). Revegetation to combat tree decline in the midlands and derwent valley lowlands of Tasmania: Practices for improved plant establishment. Ecological Management & Restoration, 4(1), 29-36.
Coates, D. J., & Atkins, K. A. (2001). Priority setting and the conservation of Western Australia's diverse and highly endemic flora. Biological Conservation, 97(2), 251-263.
References
141
Cochrane, A., Monks, L.T., & Juszkiewicz, S. (2000). Translocations trials for four threatened Western Australian plant taxa. Danthonia, 9(3), 7-9.
Cochrane, A., Crawford, A. D., & Monks, L. T. (2007). The significance of ex situ seed conservation to reintroduction of threatened plants. Australian Journal of Botany, 55(3), 356-361.
Cohen, Y., Alchanatis, V., Meron, M., Saranga, Y., & Tsipris, J. (2005). Estimation of leaf water potential by thermal imagery and spatial analysis. Journal of Experimental Botany, 56(417), 1843-1852.
Cohn, J. S., & Bradstock, R. A. (2000). Factors affecting post-fire seedling establishment of selected mallee understorey species. Australian Journal of Botany, 48(1), 59-70.
Collinge, S. K., & Louda, S. M. (1988). Herbivory by leaf miners in response to experimental shading of a native crucifer. Oecologia, 75(4), 559-566.
Cooke, S. J., Sack, L., Franklin, C. E., Farrell, A. P., Beardall, J., Wikelski, M., & Chown, S. L. (2013). What is conservation physiology? Perspectives on an increasingly integrated and essential science. Conservation Physiology, 1(1).
Cooke, S. J., & Suski, C. D. (2008). Ecological restoration and physiology: An overdue integration. Bioscience, 58(10), 957-968.
Cowling, R. M., Rundel, P. W., Lamont, B. B., Arroyo, M. K., & Arianoutsou, M. (1996). Plant diversity in mediterranean-climate regions. Trends in Ecology and Evolution, 11(9), 362-366.
Crombie, D., Tippett, J., & Hill, T. (1988). Dawn water potential and root depth of trees and understorey species in southwestern Australia. Australian Journal of Botany, 36(6), 621-631.
CSIRO. (2007). Climate change in Australia - Technical Report 2007. Melbourne, Victoria, Australia: CSIRO Bureau of Meteorology.
Cuesta, B., Villar-Salvador, P., Puértolas, J., Jacobs, D. F., & Rey Benayas, J. M. (2010). Why do large, nitrogen rich seedlings better resist stressful transplanting conditions? A physiological analysis in two functionally contrasting Mediterranean forest species. Forest Ecology and Management, 260(1), 71-78.
Dalrymple, S. E., Banks, E., Stewart, G. B., & Pullin, A. S. (2012). A meta-analysis of threatened plant reintroductions from across the globe. In J. Maschinski & K. E. Haskins (Eds.), Plant Reintroduction in a Changing Climate: Promises and Perils. Washington, DC: Island Press.
Davidson, D. W., & Morton, S. R. (1984). Dispersal adaptations of some Acacia species in the Australian arid zone. Ecology, 65(4), 1038-1051.
Davis, F. W., Borchert, M. I., & Odion, D. C. (1989). Establishment of microscale vegetation pattern in maritime chaparral after fire. Plant Ecology, 84(1), 53-67-67.
Davis, S. D., & Mooney, H. A. (1986). Tissue water relations of four co-occurring chaparral shrubs. Oecologia, 70(4), 527-535.
De'ath, G., & Fabricius, K. E. (2000). Classification and regression trees: A powerful yet simple technique for ecological data analysis. Ecology, 81(11), 3178-3192.
del Cacho, M., & Lloret, F. (2012). Resilience of Mediterranean shrubland to a severe drought episode: The role of seed bank and seedling emergence. Plant Biology, 14(3), 458-466.
References
142
Department of Conservation and Land Management. (2004). Seedling survival, soil
seed bank status and disturbance response of the Endangered Dryandra ionthocarpa A.S George ssp. ionthocarpa ms (Proteaceae), Albany, Western Australia
Department of Conservation and Land Management. (2005). Kamballup Dryandra (Dryandra ionthocarpa subsp ionthocarpa ms) Interim recovery plan, Albany, Western Australia
Department of Environment and Conservation. (2008). Approved conservation advice for Acacia awestoniana (Stirling range wattle): Department of Environment and Conservation, Western Australia.
Department of Environment and Conservation. (2008). Banksia ionthocarpa subsp. ionthocarpa recovery plan: Department of Environment and Conservation, Western Australia.
Department of Environment and Conservation. (2009). Translocation proposal: Acacia awestoniana- Stirling Range Wattle, Department of Environment and Conservation, Western Australia
Department of Parks and Wildlife. (2013). Threatened flora rankings, Department of Parks and Wildlife, Western Australia
Devine, W. D., Harrington, C. A., & Leonard, L. P. (2007). Post-planting treatments increase growth of oregon white oak (Quercus garryana Dougl. Ex Hook.) seedlings. Restoration Ecology, 15(2), 212-222.
Dixon, R. K., Pallardy, S. G., Garrett, H. E., Cox, G. S., & Sander, I. L. (1983). Comparative water relations of container-grown and bare-root ectomycorrhizal and nonmycorrhizal Quercus velutina seedlings. Canadian Journal of Botany, 61(6), 1559-1565.
Dodd, J., & Bell, D. T. (1993). Water relations of the canopy species in a Banksia woodland, Swan coastal plain, Western Australia. Australian Journal of Ecology, 18(3), 281-293.
Dodd, J., Heddle, E., Pate, J., & Dixon, K. (1984). Rooting patterns of sandplain plants and their functional significance. In J. S. Beard & J. S. Pate (Eds.), Kwongan, Plant Life of the Sandplain (pp. 146-177). Western Australia: University of Western Australia Press.
Donovan, L., & Ehleringer, J. (1991). Ecophysiological differences among juvenile and reproductive plants of several woody species. Oecologia, 86(4), 594-597.
Drayton, B., & Primack, R. B. (2000). Rates of success in the reintroduction by four methods of several perennial plant species in eastern Massachusetts. Rhodora, 102(911), 299-331.
Drayton, B., & Primack, R. B. (2012). Success rates for reintroductions of eight perennial plant species after 15 years. Restoration Ecology, 20(3), 299-303.
Eldridge, D. J., Westoby, M., & Holbrook, K. G. (1991). Soil-surface characteristics, microtopography and proximity to mature shrubs: Effects on survival of several cohorts of Atriplex vesicaria seedlings. Journal of Ecology, 79(2), 357-364.
Ellstrand, N. C., & Elam, D. R. (1993). Population genetic consequences of small population size: Implications for plant conservation. Annual Review of Ecology and Systematics, 24(1), 217-242.
References
143
Enquist, B. J., & Niklas, K. J. (2002). Global allocation rules for patterns of biomass partitioning in seed plants. Science, 295(5559), 1517-1520.
Enright, N. J., & Lamont, B., B. (1992). Survival, growth and water relations of Banksia seedlings on a sand mine rehabilitation site and adjacent scrub-heath sites. Journal of Applied Ecology, 29(3), 663-671.
Facelli, J., & Pickett, S. (1991). Plant litter: Its dynamics and effects on plant community structure. The Botanical Review, 57(1), 1-32.
Fahrig, L. (2003). Effects of habitat fragmentation on biodiversity. Annual Review of Ecology, Evolution, and Systematics, 34, 487-515.
Falster, D. S., & Westoby, M. (2005). Tradeoffs between height growth rate, stem persistence and maximum height among plant species in a post-fire succession. Oikos, 111(1), 57-66.
Fitzpatrick, M. C., Gove, A. D., Sanders, N. J., & Dunn, R. R. (2008). Climate change, plant migration, and range collapse in a global biodiversity hotspot: The Banksia (Proteaceae) of Western Australia. Global Change Biology, 14(6), 1337-1352.
Fowler, N. L. (1988). What is a safe site?: Neighbor, litter, germination date, and patch effects. Ecology, 69(4), 947-961.
Franzese, J., Ghermandi, L., & Donaldo, B. (2009). Post-fire shrub recruitment in a semi-arid grassland: The role of microsites. Journal of Vegetation Science, 20(2), 251-259.
García, D., & Ramón Obeso, J. (2003). Facilitation by herbivore-mediated nurse plants in a threatened tree, Taxus baccata: Local effects and landscape level consistency. Ecography, 26(6), 739-750.
Garrido, J., Rey, P., & Herrera, C. (2007). Regional and local variation in seedling emergence, mortality and recruitment of a perennial herb in Mediterranean mountain habitats. Plant Ecology, 190(1), 109-121.
Giam, X., Bradshaw, C. J. A., Tan, H. T. W., & Sodhi, N. S. (2010). Future habitat loss and the conservation of plant biodiversity. Biological Conservation, 143(7), 1594-1602.
Gilbert, G. S., Harms, K. E., Hamill, D. N., & Hubbell, S. P. (2001). Effects of seedling size, El Niño drought, seedling density, and distance to nearest conspecific adult on 6-year survival of Ocotea whitei seedlings in panamá. Oecologia, 127(4), 509-516.
Gillespie, I. G., & Allen, E. B. (2004). Fire and competition in a southern california grassland: Impacts on the rare forb Erodium macrophyllum. Journal of Applied Ecology, 41(4), 643-652.
Godefroid, S., Piazza, C., Rossi, G., Buord, S., Stevens, A.-D., Aguraiuja, R., Cowell, C., Weekley, C. W., Vogg, G., Iriondo, J. M., Johnson, I., Dixon, B., Gordon, D., Magnanon, S., Valentin, B., Bjureke, K., Koopman, R., Vicens, M., Virevaire, M., & Vanderborght, T. (2011). How successful are plant species reintroductions? Biological Conservation, 144(2), 672-682.
Gomez-Aparicio, L. (2008). Spatial patterns of recruitment in Mediterranean plant species: Linking the fate of seeds, seedlings and saplings in heterogeneous landscapes at different scales. Journal of Ecology, 96(6), 1128-1140.
GÓmez-Aparicio, L., GÓmez, J. M., & Zamora, R. (2005). Microhabitats shift rank in suitability for seedling establishment depending on habitat type and climate. Journal of Ecology, 93(6), 1194-1202.
References
144
Gómez-Aparicio, L., Zamora, R., Gómez, J. M., Hódar, J. A., Castro, J., & Baraza, E. (2004). Applying plant facilitation to forest restoration: A meta-analysis of the use of shrubs as nurse plants. Ecological Applications, 14(4), 1128-1138.
González-Orozco, C. E., Laffan, S. W., & Miller, J. T. (2011). Spatial distribution of species richness and endemism of the genus Acacia in Australia. Australian Journal of Botany, 59(7), 601-609.
Gordon, D. R. (1996). Experimental translocation of the endangered shrub Apalachicola rosemary Conradina glabra to the Apalachicola bluffs and ravines preserve, Florida. Biological Conservation, 77(1), 19-26.
Grant, O. M., Tronina, Ł., Jones, H. G., & Chaves, M. M. (2007). Exploring thermal imaging variables for the detection of stress responses in grapevine under different irrigation regimes. Journal of Experimental Botany, 58(4), 815-825.
Green, J. J., Baddeley, J. A., Cortina, J., & Watson, C. A. (2005). Root development in the Mediterranean shrub Pistacia lentiscus as affected by nursery treatments. Journal of Arid Environments, 61(1), 1-12.
Grieve, J. R., & Hellmuth, E. O. (1970). Eco-physiology of Western Australian plants. Oecologia Plantarium, 5, 33-68.
Groom, P. K., Froend, R. H., & Mattiske, E. M. (2000). Impact of groundwater abstraction on a Banksia woodland, Swan coastal plain, Western Australia. Ecological Management & Restoration, 1(2), 117-124.
Groom, P. K. (2004). Rooting depth and plant water relations explain species distribution patterns within a sandplain landscape. Functional Plant Biology, 31(5), 423-428.
Grossnickle, S. (2005). Importance of root growth in overcoming planting stress. New Forests, 30(2-3), 273-294.
Grubb, P. J. (1977). The maintenance of species-richness in plant communities: The importance of the regeneration niche. Biological Reviews, 55, 107-145.
Guerrant, E. O., & Kaye, T. N. (2007). Reintroduction of rare and endangered plants: Common factors, questions and approaches. Australian Journal of Botany, 55, 362-370.
Guerrant, E. O. (2012). Characterizing two decades of rare plant reintroductions. In J. Maschinski & K. E. Haskins (Eds.), Plant Reintroductions in a Changing Climate: Promises and Perils. Washington, DC: Island Press.
Guerrant, E. O. (2013). The value and propriety of reintroduction as a conservation tool for rare plants. Botany, 91(5), v-x.
Hallett, L., Standish, R., & Hobbs, R. (2011). Seed mass and summer drought survival in a Mediterranean-climate ecosystem. Plant Ecology, 212(9), 1479-1489.
Hammill, K. A., Bradstock, R. A., & Allaway, W. G. (1998). Post-fire seed dispersal and species re-establishment in Proteaceous heath. Australian Journal of Botany, 46(4), 407-419.
Hanley, M. E., & Lamont, B. B. (2001). Herbivory, serotiny and seedling defence in Western Australian Proteaceae. Oecologia, 126(3), 409-417.
Harper, J. L. (1977). Population Biology of Plants. London: Academic Press. Harris, S., Arnall, S., Byrne, M., Coates, D., Hayward, M., Martin, T., Mitchell, N., &
Garnett, S. (2013). Whose backyard? Some precautions in choosing recipient sites for assisted colonisation of Australian plants and animals. Ecological Management & Restoration, 14(2), 106-111.
References
145
Hennessy, K., Fawcett, R., Kironoa, D., Mpelasoka, F., Jones, D., Bathols, J., Whetton, P., Stafford Smith, M., Howden, M., Mitchell, C., & Plummer, N. (2008). An assessment of the impact of climate change on the nature and frequency of exceptional climatic events Canberra, Australia: CSIRO Bureau of Meteorology.
Hernández, E., Vilagrosa, A., Pausas, J., & Bellot, J. (2010). Morphological traits and water use strategies in seedlings of Mediterranean coexisting species. Plant Ecology, 207(2), 233-244.
Hnatiuk, R., & Hopkins, A. (1980). Western Australian species-rich Kwongan (sclerophyllous shrubland) affected by drought. Australian Journal of Botany, 28(6), 573-585.
Hobbs, R. J. (2001). Synergisms among habitat fragmentation, livestock grazing, and biotic invasions in southwestern Australia. Conservation Biology, 15(6), 1522-1528.
Hocking, M. (2011). Establishment of the threatened translocated species Acacia awestoniana and Banksia ionthocarpa subsp. ionthocarpa was enhanced by soil nutrients and larger initial sized seedlings. Unpublished Honours, University of Western Australia, Perth, Western Australia.
Hoffmann, M., Brooks, T., Fonseca, G., Gascon, C., Hawkins, A., James, R., Langhammer, P., Mittermeier, R., Pilgrim, J., Rodrigues, A., & Silva, J. (2008). Conservation planning and the IUCN red list. Endangered Species Research, 6(2), 113-125.
Hofmeister, P. (2010). Representational complexity and memory retrieval in language comprehension. Language and Cognitive Processes, 26(3), 376-405.
Holl, K. D., & Hayes, G. F. (2006). Challenges to introducing and managing disturbance regimes for Holocarpha macradenia, an endangered annual grassland forb. Conservation Biology, 20(4), 1121-1131.
Holmgren, M., Segura, A., & Fuentes, E. (2000). Limiting mechanisms in the regeneration of the chilean matorral – experiments on seedling establishment in burned and cleared mesic sites. Plant Ecology, 147(1), 49-57.
Hopper, S.D, & Maslin, B. (1978). Phytogeography of Acacia in Western Australia. Australian Journal of Botany, 26(1), 63-78.
Hopper, S. D. (1979). Biogeographical aspects of speciation in the southwest Australian flora. Annual Review of Ecology and Systematics, 10(1), 399-422.
Hopper, S. D. (2009). Ocbil theory: Towards an integrated understanding of the evolution, ecology and conservation of biodiversity on old, climatically buffered, infertile landscapes. Plant and Soil, 322, 49-86.
Hopper, S. D., & Gioia, P. (2004). The southwest Australian floristic region: Evolution and conservation of a global hot spot of biodiversity. Annual Review of Ecology, Evolution and Systematics, 35, 623-650.
Hsiao, T. C., Acevedo, E., Fereres, E., & Henderson, D. W. (1976). Water stress, growth, and osmotic adjustment. Philosophical Transactions of the Royal Society of London. B, Biological Sciences, 273(927), 479-500.
Huggins, T. R., Prigge, B. A., Sharifi, M. R., & Rundel, P. W. (2010). The effects of long-term drought on host plant canopy condition and survival of the endangered Astragalus jaegerianus (Fabaceae). Madroño, 57(2), 120-128.
Hughes, L. (2003). Climate change and Australia: Trends, projections and impacts. Austral Ecology, 28(4), 423-443.
References
146
Ibáñez, I., & Schupp, E. W. (2001). Positive and negative interactions between environmental conditions affecting Cercocarpus ledifolius seedling survival. Oecologia, 129(4), 543-550.
Ibáñez, I., & Schupp, E. W. (2002). Effects of litter, soil surface conditions, and microhabitat on Cercocarpus ledifolius Nutt seedling emergence and establishment. Journal of Arid Environments, 52(2), 209-221.
IUCN. (1998). Guidelines for Re-introductions. Gland, Switzerland and Cambridge, UK: IUCN; SSC Re-introduction Specialist Group.
IUCN. (2012). IUCN Red List Categories and Criteria, . Gland, Switzerland and Cambridge, UK.
IUCN. ( 2013). The IUCN Red List of Threatened Species version 2013.2. Retrieved 8th August 2013 from http://www.iucnredlist.org.
Jones, A. S., Lamont, B. B., Fairbanks, M. M., & Rafferty, C. M. (2003). Kangaroos avoid eating seedlings with or near others with volatile essential oils. Journal of Chemical Ecology, 29(12), 2621-2635.
Jones, H. G. (1999). Use of thermography for quantitative studies of spatial and temporal variation of stomatal conductance over leaf surfaces. Plant, Cell & Environment, 22(9), 1043-1055.
Jones, H. G. (2004). Application of thermal imaging and infrared sensing in plant physiology and ecophysiology. In J. A. Callow (Ed.), Advances in Botanical Research (Vol. 41, pp. 107-163): Academic Press.
Jones, H. G., Stoll, M., Santos, T., Sousa, C., Chaves, M. M., & Grant, O. M. (2002). Use of infrared thermography for monitoring stomatal closure in the field: Application to grapevine. Journal of Experimental Botany, 53(378), 2249-2260.
Josse, J., Chavent, M., Liquet, B., & Husson, F. (2012). Handling missing values with regularized iterative multiple correspondence analysis. Journal of Classification, 29(1), 91-116.
Jump, A. S., & Peñuelas, J. (2005). Running to stand still: Adaptation and the response of plants to rapid climate change. Ecology Letters, 8(9), 1010-1020.
Jumpponen, A., Vare, H., Mattson, K. G., Ohtonen, R., & Trappe, J. M. (1999). Characterization of 'safe sites' for pioneers in primary succession on recently deglaciated terrain. Journal of Ecology, 87(1), 98-105.
Jurado, E., & Westoby, M. (1992). Seedling growth in relation to seed size among species of arid Australia. Journal of Ecology, 80(3), 407-416.
Jusaitis, M. (2005). Translocation trials confirm specific factors affecting the establishment of three endangered plant species. Ecological Management and Restoration, 6(1), 61-67.
Jusaitis, M., Polomka, L., & Sorensen, B. (2004). Habitat specificity, seed germination and experimental translocation of the endangered herb brachycome muelleri (Asteraceae). Biological Conservation, 116(2), 251-266.
Kaye, T. N. (2008). Vital steps toward success of endangered plant reintroductions. Native Plants Journal, 9(3), 313-322.
Keeley, J. K., & Fotheringham, C. J. (2000). Role of fire in regeneration of seed. In M. Fenner (Ed.), Seeds: The Ecology of Regeneration in Plant Communities: CAB International.
References
147
Kennedy, K. L., Albrecht, M. A., Guerrant Jr, E. O., Dalrymple, S. E., Maschinski, J., & Haskins, K. E. (2012). Synthesis and future directions. In J. Maschinski & K. E. Haskins (Eds.), Plant Reintroduction in a Changing Climate: Promises and Perils. Washington, DC: Island Press.
Kolb, P. F., & Robberecht, R. (1996). High temperature and drought stress effects on survival of Pinus ponderosa seedlings. Tree Physiology, 16(8), 665-672.
Kooten, O., & Snel, J. F. H. (1990). The use of chlorophyll fluorescence nomenclature in plant stress physiology. Photosynthesis Research, 25(3), 147-150.
Kozlowski, T. T., & Pallardy, S. G. (2002). Acclimation and adaptive responses of woody plants to environmental stresses. Botanical Review, 68(2), 270-334.
Lamont, B.B (1985). Gradient and zonal analysis of understorey suppression by Eucalyptus wandoo. Plant Ecology, 63(2), 49-66.
Lamont, B.B, & Connell, S. W. (1996). Biogeography of Banksia in southwestern Australia. Journal of Biogeography, 23(3), 295-309.
Lamont, B.B, & Groom, P. K. ( 2013). Seeds as a source of carbon, nitrogen, and phosphorus for seedling establishment in temperate regions: A synthesis. American Journal of Plant Sciences, 4, 30-40
Lamont, B.B, Witkowski, E. T. F., & Enright, N. J. (1993). Post-fire litter microsites: Safe for seeds, unsafe for seedlings. Ecology, 74(2), 501-512.
Lamont, B. B., & Bergl, S. M. (1991). Water relations, shoot and root architecture, and phenology of three co-occurring Banksia species: No evidence for niche differentiation in the pattern of water use. Oikos, 60(3), 291-298.
Lamont, B. B., Enright, N. J., Witkowski, E. T. F., & Groeneveld, J. (2007). Conservation biology of Banksias: Insights from natural history to simulation modelling. Australian Journal of Botany, 55(3), 280-292.
Lamont, B. B., & Groom, P. K. (1998). Seed and seedling biology of the woody-fruited Proteaceae. Australian Journal of Botany, 46(4), 387-406.
Lamont, B. B., Groom, P. K., Richards, M. B., & Witkowski, E. T. F. (1999). Recovery of Banksia and Hakea communities after fire in Mediterranean Australia -- the role of species identity and functional attributes. Diversity and Distributions, 5(1/2), 15-26.
Lawrence, B. A., & Kaye, T. N. (2009). Reintroduction of Castilleja levisecta: Effects of ecological similarity, source population genetics, and habitat quality. Restoration Ecology, 19(2), 166-176.
Leinonen, I., Grant, O. M., Tagliavia, C. P. P., Chaves, M. M., & Jones, H. G. (2006). Estimating stomatal conductance with thermal imagery. Plant, Cell & Environment, 29(8), 1508-1518.
Lenthe, J. H., Oerke, E. C., & Dehne, H. W. (2007). Digital infrared thermography for monitoring canopy health of wheat. Precision Agriculture, 8(1-2), 15-26.
Lilley, J. M., & Ludlow, M. M. (1996). Expression of osmotic adjustment and dehydration tolerance in diverse rice lines. Field Crops Research, 48(2–3), 185-197.
Llorens, T. M., Byrne, M., Yates, C. J., Nistelberger, H. M., & Coates, D. J. (2012). Evaluating the influence of different aspects of habitat fragmentation on mating patterns and pollen dispersal in the bird-pollinated Banksia sphaerocarpa var caesia. Molecular Ecology, 21(2), 314-328.
References
148
Lloret, F., Peñuelas, J., & Estiarte, M. (2004). Experimental evidence of reduced diversity of seedlings due to climate modification in a Mediterranean-type community. Global Change Biology, 10(2), 248-258.
Lloret, F., Peñuelas, J., Estiarte, M., & Ezcurra, E. (2005). Effects of vegetation canopy and climate on seedling establishment in Mediterranean shrubland. Journal of Vegetation Science, 16(1), 67-76.
Ludwig, J. A., Reynolds, J. F., & Whitson, P. D. (1975). Size-biomass relationships of several chihuahuan desert shrubs. American Midland Naturalist, 94(2), 451-461.
Maestre, F. T., Cortina, J., Bautista, S., Bellot, J., & Vallejo, R. (2003). Small-scale environmental heterogeneity and spatiotemporal dynamics of seedling establishment in a semiarid degraded ecosystem. Ecosystems, 6(7), 630-643.
Maiorana, V. C. (1981). Herbivory in sun and shade. Biological Journal of the Linnean Society, 15(2), 151-156.
Malcolm, J. R., Liu, C., Neilson, R. P., Hansen, L., & Hannah, L. E. E. (2006). Global warming and extinctions of endemic species from biodiversity hotspots. Conservation Biology, 20(2), 538-548.
Maliva, R., & Missimer, T. (2012). Aridity and drought. In Arid Lands Water Evaluation and Management (pp. 21-39): Springer Berlin Heidelberg.
Mappin, K. A., Pate, J. S., & Bell, T. L. (2003). Productivity and water relations of burnt and long-unburnt semi-arid shrubland in Western Australia. Plant and Soil, 257(2), 321-340.
Maschinski, J., Baggs, J. E., & Sacchi, C. F. (2004). Seedling recruitment and survival of an endangered limestone endemic in its natural habitat and experimental reintroduction sites. American Journal of Botany, 91(5), 689-698.
Maschinski, J., & Duquesnel, J. (2007). Successful reintroductions of the endangered long-lived Sargent's Cherry Palm, Pseudophoenix sargentii, in the Florida keys. Biological Conservation, 134(1), 122-129.
Maschinski, J., Falk, D. A., Wright, S., Possley, J., Roncal, J., & Wendelberger, K. S. (2012). Optimal locations for plant reintroduction in a changing world. In J. Maschinski & K. E. Haskins (Eds.), Plant Reintroduction in a Changing Climate: Promises and Perils. Washington, DC: Island Press.
Matusick, G., Ruthrof, K., Brouwers, N., Dell, B., & Hardy, G. J. (2013). Sudden forest canopy collapse corresponding with extreme drought and heat in a mediterranean-type eucalypt forest in southwestern Australia. European Journal of Forest Research, 132(3), 497-510.
Maunder, M. (1992). Plant reintroduction: An overview. Biodiversity and Conservation, 1(1), 51-61.
McDowell, N. G, Pockman, W. T., Allen, C. D., Breshears, D. D., Cobb, N., Kolb, T., Plaut, J., Sperry, J., West, A., Williams, D. G., & Yepez, E. A. (2008). Mechanisms of plant survival and mortality during drought: Why do some plants survive while others succumb to drought? New Phytologist, 178(4), 719-739.
References
149
McDowell, N. G., Fisher, R. A., Xu, C., Domec, J. C., Hölttä, T., Mackay, D. S., Sperry, J. S., Boutz, A., Dickman, L., Gehres, N., Limousin, J. M., Macalady, A., Martínez-Vilalta, J., Mencuccini, M., Plaut, J. A., Ogée, J., Pangle, R. E., Rasse, D. P., Ryan, M. G., Sevanto, S., Waring, R. H., Williams, A. P., Yepez, E. A., & Pockman, W. T. (2013). Evaluating theories of drought-induced vegetation mortality using a multimodel–experiment framework. New Phytologist, 200(2), 304-321.
McIntyre, S. (2011). Ecological and anthropomorphic factors permitting low-risk assisted colonization in temperate grassy woodlands. Biological Conservation, 144(6), 1781-1789.
Mediavilla, S., & Escudero, A. (2004). Stomatal responses to drought of mature trees and seedlings of two co-occurring Mediterranean oaks. Forest Ecology and Management, 187(2–3), 281-294.
Mellin, C., Bradshaw, C. J. A., Meekan, M. G., & Caley, M. J. (2010). Environmental and spatial predictors of species richness and abundance in coral reef fishes. Global Ecology and Biogeography, 19(2), 212-222.
Méndez, M., García, D., Maestre, F. T., & Escudero, A. (2008). More ecology is needed to restore Mediterranean ecosystems: A reply to Valladares and Gianoli. Restoration Ecology, 16(2), 210-216.
Menges, E. S. (2008). Restoration demography and genetics of plants: When is translocation successful? Australian Journal of Botany, 56, 187-196.
Merwin, L., He, T., & Lamont, B. B. (2012). Phylogenetic and phenotypic structure among Banksia communities in south-western Australia. Journal of Biogeography, 39(2), 397-407.
Millar, M. A., Byrne, M., & Coates, D. J. (2010). The maintenance of disparate levels of clonality, genetic diversity and genetic differentiation in disjunct subspecies of the rare Banksia ionthocarpa. Molecular Ecology, 19(19), 4217-4227.
Mitchell, P., Veneklaas, E., Lambers, H., & Burgess, S. O. (2008). Using multiple trait associations to define hydraulic functional types in plant communities of south-western Australia. Oecologia, 158(3), 385-397.
Mittermeier, R., Turner, W., Larsen, F., Brooks, T., & Gascon, C. (2011). Global biodiversity conservation: The critical role of hotspots. In F. E. Zachos & J. C. Habel (Eds.), Biodiversity hotspots (pp. 3-22): Springer Berlin Heidelberg.
Moles, A. T., & Westoby, M. (2004). What do seedlings die from and what are the implications for evolution of seed size? Oikos, 106(1), 193-199.
Moller, M., Alchanatis, V., Cohen, Y., Meron, M., Tsipris, J., Naor, A., Ostrovsky, V., Sprintsin, M., & Cohen, S. (2007). Use of thermal and visible imagery for estimating crop water status of irrigated grapevine. Journal of Experimental Botany, 58(4), 827-838.
Monks, L. (2008). Experimental approaches in threatened plant translocations: How failures can still lead to success. Australasian Plant Conservation, 17(3), 8-10.
Monks, L., Coates, D., Bell, T., & Bowles, M. (2012). Determining success criteria for reintroductions of threatened long-lived plants. In J. Maschinski & K. E. Haskins (Eds.), Plant reintroduction in a changing climate: Promises and Perils. Washington, DC: Island Press.
Monks, L. T. (1999). Conservation biology of the rare and threatened Dryandra ionthocarpa, D. mimica and D. serra. Curtin University of Technology, Perth.
References
150
Monks, L. T., & Coates, D. (2002). The translocation of two critically endangered Acacia species. Conservation Science Western Australia, 4(3), 54-61.
Montalvo, A. M., & Ellstrand, N. C. (2000). Transplantation of the subshrub Lotus scoparius: Testing the home-site advantage hypothesis. Conservation Biology, 14(4), 1034-1045.
Morgan, J. (1992). Osmotic components and properties associated with genotypic differences in osmoregulation in wheat. Functional Plant Biology, 19(1), 67-76.
Morgan, J. M. (1984). Osmoregulation and water stress in higher plants. Annual Review of Plant Physiology, 35(1), 299-319.
Morgan, J. W. (1999). Have tubestock plantings successfully established populations of rare grassland species into reintroduction sites in western Victoria? Biological Conservation, 89(3), 235-243.
Morgan, J. W. (2003). Plants, osmotic adjustment. In B. A. Stewart & T. A. Howell (Eds.), Encyclopedia of water science. New York: Marcel Dekker, Inc.
Muller, C. H., Hanawalt, R. B., & McPherson, J. K. (1968). Allelopathic control of herb growth in the fire cycle of california chaparral. Bulletin of the Torrey Botanical Club, 95(3), 225-231.
Muth, N. Z., Kluger, E. C., Levy, J. H., Edwards, M. J., & Niesenbaum, R. A. (2008). Increased per capita herbivory in the shade: Necessity, feedback, or luxury consumption. Ecoscience, 15(2), 182-188.
Myers, N., Mittermeier, R. A., Mittermeier, C. G., da Fonseca, G. A. B., & Kent, J. (2000). Biodiversity hotspots for conservation priorities. Nature, 403(6772), 853-858.
Nardini, A., Lo Gullo, M. A., Trifilò, P., & Salleo, S. (2013). The challenge of the Mediterranean climate to plant hydraulics: Responses and adaptations. Environmental and Experimental Botany, (in press.).
Nicotra, A. B., Atkin, O. K., Bonser, S. P., Davidson, A. M., Finnegan, E. J., Mathesius, U., Poot, P., Purugganan, M. D., Richards, C. L., Valladares, F., & van Kleunen, M. (2010). Plant phenotypic plasticity in a changing climate. Trends in Plant Science, 15(12), 684-692.
Niinemets, Ü. (2001). Global-scale climatic controls of leaf dry mass per area, density, and thickness in trees and shrubs. Ecology, 82(2), 453-469.
Niinemets, Ü. (2010). Responses of forest trees to single and multiple environmental stresses from seedlings to mature plants: Past stress history, stress interactions, tolerance and acclimation. Forest Ecology and Management, 260(10), 1623-1639.
Nilsson, U., & Örlander, G. (1995). Effects of regeneration methods on drought damage to newly planted Norway spruce seedlings. Canadian Journal of Forest Research, 25(5), 790-802.
Oerke, E.-C., Steiner, U., Dehne, H.-W., & Lindenthal, M. (2006). Thermal imaging of cucumber leaves affected by downy mildew and environmental conditions. Journal of Experimental Botany, 57(9), 2121-2132.
Otieno, D. O., Kinyamario, J. I., & Omenda, T. O. (2001). Growth features of Acacia tortilis and Acacia xanthophloea seedlings and their response to cyclic soil drought stress. African Journal of Ecology, 39(2), 126-132.
Padilla, F.M., Miranda, J. d. D., & Pugnaire, F. (2007). Early root growth plasticity in seedlings of three Mediterranean woody species. Plant and Soil, 296(1-2), 103-113.
References
151
Padilla, F. M., & Pugnaire, F. I. (2006). The role of nurse plants in the restoration of degraded environments. Frontiers in Ecology and the Environment, 4(4), 196-202.
Padilla, F. M., & Pugnaire, F. I. (2007). Rooting depth and soil moisture control Mediterranean woody seedling survival during drought. Functional Ecology, 21(3), 489-495.
Pansu, M., & Gautheyrou, J. (2006). Handbook of Soil Analysis: Mineralogical, Organic and Inorganic Methods. The Netherlands: Springer.
Parsons, M., Koch, J., Lamont, B., Vlahos, S., & Fairbanks, M. (2006). Planting density effects and selective herbivory by kangaroos on species used in restoring forest communities. Forest Ecology and Management, 229, 39 - 49.
Parsons, M., Rafferty, C., Lamont, B., Dods, K., & Fairbanks, M. (2007). Relative effects of mammal herbivory and plant spacing on seedling recruitment following fire and mining. BMC Ecology, 7(1), 13.
Pasquet-Kok, J., Creese, C., & Sack, L. (2010). Turning over a new ‘leaf’: Multiple functional significances of leaves versus phyllodes in Hawaiian Acacia koa. Plant, Cell & Environment, 33(12), 2084-2100.
Pate, J. S., & Bell, T. L. (1999). Application of the ecosystem mimic concept to the species-rich Banksia woodlands of Western Australia. Agroforestry Systems, 45(1-3), 303-341.
Pate, J. S., Froend, R. H., Bowen, B. J., Hansen, A., & Kuo, J. (1990). Seedling growth and storage characteristics of seeder and resprouter species of Mediterranean-type ecosystems of S.W. Australia Annals of Botany, 65(6), 585-601.
Pavlik, B. M., Nickrent, D. L., & Howald, A. M. (1993). The recovery of an endangered plant. I. Creating a new population of Amsinckia grandiflora. Conservation Biology, 7(3), 510-526.
Peterson, C. L., Kaufmann, G. S., Vandello, C., & Richardson, M. L. (2013). Parent genotype and environmental factors influence introduction success of the critically endangered Savannas mint (Dicerandra immaculata var. savannarum). PLoS ONE, 8(4), 1-7.
Petru, M., & Menges, E. S. (2003). Seedling establishment in natural and experimental Florida scrub gaps. Journal of the Torrey Botanical Society, 130(2), 89-100.
Pohlman, C. L., Nicotra, A. B., & Murray, B. R. (2005). Geographic range size, seedling ecophysiology and phenotypic plasticity in Australian Acacia species. Journal of Biogeography, 32(2), 341-351.
Poorter, H., Niinemets, Ü., Poorter, L., Wright, I. J., & Villar, R. (2009). Causes and consequences of variation in leaf mass per area (LMA): A meta-analysis. New Phytologist, 182(3), 565-588.
Poorter, H., Niklas, K. J., Reich, P. B., Oleksyn, J., Poot, P., & Mommer, L. (2012). Biomass allocation to leaves, stems and roots: Meta-analyses of interspecific variation and environmental control. New Phytologist, 193(1), 30-50.
Poot, P., Hopper, S. D., & van Diggelen, J. M. H. (2012). Exploring rock fissures: Does a specialized root morphology explain endemism on granite outcrops? Annals of Botany, 110(2), 291-300.
Poot, P., & Veneklaas, E. (2013). Species distribution and crown decline are associated with contrasting water relations in four common sympatric eucalypt species in southwestern Australia. Plant and Soil, 364(1-2), 409-423.
References
152
Pressey, R. L., Cabeza, M., Watts, M. E., Cowling, R. M., & Wilson, K. A. (2007). Conservation planning in a changing world. Trends in Ecology & Evolution, 22(11), 583-592.
Primack, R. P. (1996). Lessons from ecological theory: Dispersal, establishment and population structure. In D. A. Falk, C. I. Millar & M. Olwell (Eds.), Restoring diversity: Strategies for Reintroduction of Endangered Plants (pp. 209-233). Washington, DC: Island Press.
Quero, J., Sterck, F., Martínez-Vilalta, J., & Villar, R. (2011). Water-use strategies of six co-existing Mediterranean woody species during a summer drought. Oecologia, 166(1), 45-57.
R Core Team. (2013). R: A language and environment for statistical computing. Vienna, Austria,http://www.R-project.org/
Raabová, J., Münzbergová, Z., & Fischer, M. (2007). Ecological rather than geographic or genetic distance affects local adaptation of the rare perennial herb, Aster amellus. Biological Conservation, 139(3-4), 348-357.
Rafferty, C., Byron, B. L., & Hanley, M. E. (2005). Selective feeding by kangaroos (Macropus fuliginosus) on seedlings of Hakea species: Effects of chemical and physical defences. Plant Ecology, 177(2), 201-208.
Raulings, E., Boon, P., Bailey, P., Roache, M., Morris, K., & Robinson, R. (2007). Rehabilitation of swamp paperbark (Melaleuca ericifolia) wetlands in south-eastern Australia: Effects of hydrology, microtopography, plant age and planting technique on the success of community-based revegetation trials. Wetlands Ecology and Management, 15(3), 175-188.
Reader, R. J., Jalili, A., Grime, J. P., Spencer, R. E., & Matthews, N. (1993). A comparative study of plasticity in seedling rooting depth in drying soil. Journal of Ecology, 81(3), 543-550.
Regan, H. M., Syphard, A. D., Franklin, J., Swab, R. M., Markovchick, L., Flint, A. L., Flint, L. E., & Zedler, P. H. (2012). Evaluation of assisted colonization strategies under global change for a rare, fire-dependent plant. Global Change Biology, 18(3), 936-947.
Reich, P. B., Wright, I. J., Bares, J. C., Craine, J. M., Oleksyn, J., Westoby, M., & Walters, M. B. (2003). The evolution of plant functional variation: Traits, spectra, and strategies. International Journal of Plant Sciences, 164(S3), S143-S164.
Renton, M., Shackelford, N., & Standish, R. J. (2012). Habitat restoration will help some functional plant types persist under climate change in fragmented landscapes. Global Change Biology, 18(6), 2057-2070.
Rice, E. L. (1979). Allelopathy: An update. Botanical Review, 45(1), 15-109. Rice, K. J., Matzner, S. L., Byer, W., & Brown, J. R. (2004). Patterns of tree dieback in
Queensland, Australia: The importance of drought stress and the role of resistance to cavitation. Oecologia, 139(2), 190-198.
Rice, S. K. (1993). Vegetation establishment in post-fire Adenostoma chaparral in relation to fine-scale pattern in fire intensity and soil nutrients. Journal of Vegetation Science, 4(1), 115-124.
Ripley, B. (2013). Tree:Classification and regression trees Retrieved 16th June 2013, from http://CRAN.R-project.org/package=tree
References
153
Roncal, J., Maschinski, J., Schaffer, B., Gutierrez, S. M., & Walters, D. (2012). Testing appropriate habitat outside of historic range: The case of Amorpha herbacea var. crenulata (Fabaceae). Journal for Nature Conservation, 20(2), 109-116.
Rowland, J., & Maun, M. A. (2001). Restoration ecology of an endangered plant species: Establishment of new populations of Cirsium pitcheri. Restoration Ecology, 9(1), 60.
Ruthrof, K., Douglas, T., Calver, M., Barber, P., Dell, B., & Hardy, G. (2010). Restoration treatments improve seedling establishment in a degraded Mediterranean-type Eucalyptus ecosystem. Australian Journal of Botany, 58(8), 646-655.
Sandiford, E. M., & Barrett, S. (2010). Albany Regional Vegetation Survey: Extent, Type and Status. Western Australia Department of Environment and Conservation.
Sarrazin, F., & Barbault, R. (1996). Reintroduction: Challenges and lessons for basic ecology. Trends in Ecology & Evolution, 11(11), 474-478.
Schütz, W., Milberg, P., & Lamont, B. B. (2002). Germination requirements and seedling responses to water availability and soil type in four eucalypt species. Acta Oecologica, 23(1), 23-30.
Shea, S. R., McCormick, J., & Portlock, C. C. (1979). The effect of fires on regeneration of leguminous species in the northern jarrah (Eucalyptus marginata sm) forest of Western Australia. Australian Journal of Ecology, 4(2), 195-205.
Sheean, V. A., Manning, A. D., & Lindenmayer, D. B. (2012). An assessment of scientific approaches towards species relocations in Australia. Austral Ecology, 37(2), 204-215.
Siles, G., Rey, P. J., Alcántara, J. M., Bastida, J. M., & Herreros, J. L. (2010). Effects of soil enrichment, watering and seedling age on establishment of Mediterranean woody species. Acta Oecologica, 36(4), 357-364.
Slik, J. W. F., Paoli, G., McGuire, K., Amaral, I., Barroso, J., Bastian, M., Blanc, L., Bongers, F., Boundja, P., Clark, C., Collins, M., Dauby, G., Ding, Y., Doucet, J.-L., Eler, E., Ferreira, L., Forshed, O., Fredriksson, G., Gillet, J.-F., Harris, D., Leal, M., Laumonier, Y., Malhi, Y., Mansor, A., Martin, E., Miyamoto, K., Araujo-Murakami, A., Nagamasu, H., Nilus, R., Nurtjahya, E., Oliveira, Á., Onrizal, O., Parada-Gutierrez, A., Permana, A., Poorter, L., Poulsen, J., Ramirez-Angulo, H., Reitsma, J., Rovero, F., Rozak, A., Sheil, D., Silva-Espejo, J., Silveira, M., Spironelo, W., ter Steege, H., Stevart, T., Navarro-Aguilar, G. E., Sunderland, T., Suzuki, E., Tang, J., Theilade, I., van der Heijden, G., van Valkenburg, J., Van Do, T., Vilanova, E., Vos, V., Wich, S., Wöll, H., Yoneda, T., Zang, R., Zhang, M.-G., & Zweifel, N. (2013). Large trees drive forest aboveground biomass variation in moist lowland forests across the tropics. Global Ecology and Biogeography, 22(12), 1261-1271.
Smith, M. (2013). Threatened and Priority Flora List for Western Australia Kensington, Western Australia: Department of Parks and Wildlife.
Standish, R. J., Fontaine, J. B., Harris, R. J., Stock, W. D., & Hobbs, R. J. (2012). Interactive effects of altered rainfall and simulated nitrogen deposition on seedling establishment in a global biodiversity hotspot. Oikos, 121(12), 2014-2025.
Thomas, C. D. (2011). Translocation of species, climate change, and the end of trying to recreate past ecological communities. Trends in Ecology & Evolution, 26(5), 216-221.
References
154
Turner, N. (1981). Techniques and experimental approaches for the measurement of plant water status. Plant and Soil, 58(1), 339-366.
Turner, R. M., Alcorn, S. M., Olin, G., & Booth, J. A. (1966). The influence of shade, soil, and water on saguaro seedling establishment. Botanical Gazette, 127(2/3), 95-102.
Tyler, C. M. (1995). Factors contributing to postfire seedling establishment in chaparral: Direct and indirect effects of fire. Journal of Ecology, 83(6), 1009-1020.
UCLA: Statistical Consulting Group. (2011). What are pseudo r-squareds? Retrieved 25th November 2013, from http://www.ats.ucla.edu/stat/mult_pkg/faq/general/Psuedo_RSquareds.htm.
Urbanska, K. M. (1997). Safe sites - interface of plant population ecology and restoration ecology. In K. M. Urbanska, N. R. Webb & P. J. Edwards (Eds.), Restoration Ecology and Sustainable Development. Cambridge, UK: Cambridge University Press.
Uytvanck, J. V., Maes, D., Vandenhaute, D., & Hoffmann, M. (2008). Restoration of woodpasture on former agricultural land: The importance of safe sites and time gaps before grazing for tree seedlings. Biological Conservation, 141(1), 78-88.
Valladares, F., & Gianoli, E. (2007). How much ecology do we need to know to restore Mediterranean ecosystems? Restoration Ecology, 15(3), 363-368.
Valladares, F., & Sánchez-Gómez, D. (2006). Ecophysiological traits associated with drought in Mediterranean tree seedlings: Individual responses versus interspecific trends in eleven species. Plant Biology, 8(5), 688-697.
Varone, L., Ribas-Carbo, M., Cardona, C., Gallé, A., Medrano, H., Gratani, L., & Flexas, J. (2012). Stomatal and non-stomatal limitations to photosynthesis in seedlings and saplings of Mediterranean species pre-conditioned and aged in nurseries: Different response to water stress. Environmental and Experimental Botany, 75(0), 235-247.
Vilagrosa, A., Bellot, J., Vallejo, V. R., & Gil-Pelegrín, E. (2003). Cavitation, stomatal conductance, and leaf dieback in seedlings of two co-occurring Mediterranean shrubs during an intense drought. Journal of Experimental Botany, 54(390), 2015-2024.
Volis, S., & Blecher, M. (2010). Quasi in situ: A bridge between ex situ and in situ conservation of plants. Biodiversity and Conservation, 19(9), 2441-2454.
Volis, S., Dorman, M., Blecher, M., Sapir, Y., & Burdeniy, L. (2011). Variation partitioning in canonical ordination reveals no effect of soil but an effect of co-occurring species on translocation success in iris atrofusca. Journal of Applied Ecology, 48(1), 265-273.
Warren, C. R., Bleby, T., & Adams, M. A. (2007). Changes in gas exchange versus leaf solutes as a means to cope with summer drought in Eucalyptus marginata. Oecologia, 154(1), 1-10.
Warwick, N. W. M., & Thukten. (2006). Water relations of phyllodinous and non-phyllodinous Acacias, with particular reference to osmotic adjustment. Physiologia Plantarum, 127(3), 393-403.
Wendelberger, K. S., Fellows, M. Q. N., & Maschinski, J. (2008). Rescue and restoration: Experimental translocation of Amorpha herbacea walter var. crenulata (Rybd.) Isley into a novel urban habitat. Restoration Ecology, 16(4), 542-552.
References
155
Whisenant, S. G., Thurow, T. L., & Maranz, S. J. (1995). Initiating autogenic restoration on shallow semiarid sites. Restoration Ecology, 3(1), 61-67.
Williams, J. E., Davis, S. D., & Portwood, K. (1997). Xylem embolism in seedlings and resprouts of Adenostoma fasciculatum after fire. Australian Journal of Botany, 45(2), 291-300.
Witkowski, E., & Lamont, B. (2006). Resilience of two Banksia species to global change: Comparing results of bioclimatic modelling, demographic and translocation studies International Journal of Biodiversity Science & Management, 2(2), 59-72.
Witkowski, E. T. F. (1991). Growth and competition between seedlings of Protea repens (L.) L. And the alien invasive, Acacia saligna (Labill.) Wendl. In relation to nutrient availability. Functional Ecology, 5(1), 101-110.
Wright, I. J., Reich, P. B., Westoby, M., Ackerly, D. D., Baruch, Z., Bongers, F., Cavender-Bares, J., Chapin, T., Cornelissen, J. H. C., Diemer, M., Flexas, J., Garnier, E., Groom, P. K., Gulias, J., Hikosaka, K., Lamont, B. B., Lee, T., Lee, W., Lusk, C., Midgley, J. J., Navas, M.-L., Niinemets, Ü., Oleksyn, J., Osada, N., Poorter, H., Poot, P., Prior, L., Pyankov, V. I., Roumet, C., Thomas, S. C., Tjoelker, M. G., Veneklaas, E. J., & Villar, R. (2004). The worldwide leaf economics spectrum. Nature, 428(6985), 821-827.
Yates, C., & Ladd, P. (2010). Using population viability analysis to predict the effect of fire on the extinction risk of an endangered shrub Verticordia fimbrilepis subsp. fimbrilepis in a fragmented landscape. Plant Ecology, 211(2), 305-319.
Yates, C. J., Hobbs, R. J., & Atkins, L. (2000). Establishment of perennial shrub and tree species in degraded Eucalyptus salmonophloia (Salmon gum) remnant woodlands: Effects of restoration treatments. Restoration Ecology, 8(2), 135-143.
Yazaki, K., Sano, Y., Fujikawa, S., Nakano, T., & Ishida, A. (2010). Response to dehydration and irrigation in invasive and native saplings: Osmotic adjustment versus leaf shedding. Tree Physiology, 30(5), 597-607.
Young, T. P. (2000). Restoration ecology and conservation biology. Biological Conservation, 92(1), 73-83.
Zunzunegui, M., Díaz Barradas, M., Ain-Lhout, F., Alvarez-Cansino, L., Esquivias, M. P., & García Novo, F. (2011). Seasonal physiological plasticity and recovery capacity after summer stress in Mediterranean scrub communities. Plant Ecology, 212(1), 127-142.
Zuur, A. F., Ieno, E. N., & Smith, G. M. (2007). Analysing Ecological Data: Statistics for Biology and Health: Springer.
Appendices
156
Appendices
Appendix 1 to 3 were submitted as supplementary material as part of Chapter 2
Appendix 1– Statistical analysis of soil moisture data using the neutron probe (adjusted count
ratio – Dec 2010 neutron counts subtracted from subsequent months) as dependent on plot
watering treatments, microhabitat, soil depth and time. Plot was used a random factor and
“no water” and “open” were used as reference levels for the categorical variables “watering”
and location to “E. wandoo” respectively. The mean of each level of soil depth (20 –60cm, 60–
100cm and 100–180cm) was compared to the grand mean of all depths. Three-way
interactions were non- significant and therefore deleted from the mixed model. Bold values
indicate p<0.05.
Variable Degrees of
freedom
P-
value
Under E. wandoo 135 0.45 20–60cm depth 998 0.03 60–100 cm depth 998 0.61 Time 4 0.56 Monthly water 135 0.01 Weekly water 135 0.02 Under E. wandoo × 20–60 cm 998 0.02 Under E. wandoo × 60–100 cm 998 0.44 20–60 cm × time 998 <0.001 60–100 cm × time 998 0.05 20–60 cm × Monthly watering 998 0.003 60–100 cm × Monthly water 998 0.26 20–60 cm × Weekly water 998 0.22 60–100 cm × Weekly water 998 0.65
Appendices
157
Appendix 2 – Results of the optimal linear mixed models with seedling survival of A.
awestoniana as a binomial response variable. All models initially contained all explanatory
variables (seedling age/ time of planting, pre-translocation substrate, location to E. wandoo,
watering and seedling initial height) and were reduced in a step-wise fashion. The variables
included here are those in the optimal models with lowest (AIC) and with significant variables
(p< 0.05). Separate models were created for each monitoring month. Percentage deviance is
included in brackets which provide a comparison between the full and optimal model.
Monitoring date Optimal model for seedling survival
December 2010 Planting age /time × initial height (p<0.001) (31.2%) January 2011 Planting age /time × initial height (p<0.001) (19.3%) March 2011 Planting age /time × initial height (p<0.001)
Planting age /time × soil substrate (p = 0.009) (16.8%) June 2011 Planting age /time × initial height (p<0.001)
Planting age /time × soil substrate (p= 0.003) (13.6%)
January 2012 Planting age /time × initial height (p<0.001) Planting age /time × soil substrate (p= 0.003) (12.5%)
March 2012 Planting age /time × initial height (p<0.001) Planting age /time × soil substrate (p= 0.001) (12.3%)
July 2012 Planting age /time × initial height (p<0.001) Planting age /time × soil substrate (p< 0.001) (12.8%)
Appendices
158
Appendix 3 – The optimal linear mixed models with seedling height (mm) as a response
variable. All models included all explanatory variables and their interactions (seedling age/
time of planting, pre-translocation substrate, location to E. wandoo, watering and seedling
initial height), which were then deleted in a step-wise manner until the optimal model was
found (lowest AIC). A separate model was created for each monitoring month. For the
categorical variables “batch”, “substrate”, “watering frequency” and “location to E. wandoo”
the levels “batch A”, “potting mix”, “no watering” and “open” were used as references
respectively. The variables included were significant in the optimal models (p<0.05). A
likelihood ratio R2 in brackets was used to analyse goodness-of-fit for each optimal model.
Pseudo- R squared’s can be used to compare full model with optimal model and cannot be
compared between models (UCLA: Statistical Consulting Group, 2011).
Monitoring date Optimal model for seedling height
Jan 2011 Batch B × sand substrate(d.f. 686, p<0.001) (LRR2 = 0.7)
March 2011 Under E. wandoo × Weekly watering × batch B (d.f. 653, p = 0.001)
Batch B × sand substrate (d.f. 653, p<0.001)
Weekly watering × batch B (d.f. 653, p<0.01)
Weekly watering (d.f. 17, p = 0.01)
Batch B (d.f. 653, p<0.001)
Sand substrate (d.f. 653, p<0.001)
June 2011 Under E. wandoo × Weekly watering × batch B (d.f. 644, p = 0.003)
Batch B × sand substrate (d.f. 644, p<0.001)
Weekly watering × batch B (d.f. 644,p = 0.01)
Weekly watering (d.f. 18, p = 0.005)
Batch B (d.f. 644, p<0.001)
Sand substrate (d.f. 644, p<0.001)
Jan 2012 Batch B × sand substrate (d.f. 639, p<0.001)
Sand substrate (d.f. 639, p<0.001)
Batch B (d.f. 639, p<0.001)
Weekly watering (d.f. 21, p = 0.005) (LRR2 = 0.46)
March 2012 Under E. wandoo × Weekly watering × batch B (d.f. 625, p = 0.007)
Batch B × sand substrate (d.f. 625, p<0.001)
Weekly watering × batch B (d.f. 625, p = 0.006)
Weekly watering (d.f. 18, p = 0.01)
Batch B (d.f. 625, p<0.001)
Sand substrate (d.f. 625, p<0.001) (LRR2 = 0.47)
July 2012 Batch B × sand substrate (d.f. 626, p<0.001)
Weekly watering × batch B (d.f. 626, p = 0.03)
Weekly watering (d.f. 21, p = 0.001)
Batch B (d.f. 626, p<0.001)
Sand substrate (d.f. 626, p<0.001) (LRR2 = 0.46)
Appendices
159
Appendix 4 and 5 were submitted as supplementary material as part of Chapter 3
Appendix 4 – Results of linear mixed models with seedling survival as a binomial response
variable and microhabitat type (bare, low, tall heath), watering frequency (monthly and
weekly) and initial seedling size at planting (centred) as explanatory variables. All models
initially contained all interactions between explanatory variables and models were
subsequently reduced in a step-wise fashion. Separate models were created for seedlings in
different substrate/ planting time combinations as well as each monitoring month. The levels
‘weekly’ and ‘bare’ were used as references to compare the factors ‘watering’ and ‘habitat’.
The variables included here are those in the optimal models with lowest (AIC) and with
significant variables (p< 0.05). Non-significant indicates that all variables in the specific model
were not significant at the 5% level. Percentage deviance is included in brackets which provide
a comparison between the full and optimal model.
Monitoring date Planting substrate / planting season treatment combinations
Potting / Early planting Potting / Late planting Sand / Early planting
Nov 2010 Centred seedling volume (p = 0.007) (5.7%)
Centred height (p = 0.005) (32.9%)
Non- significant
Jan 2011 Non- significant Tall heath (p = 0.02) Hand watering (p = 0.03) Centred seedling volume (p = 0.008) (12.4%)
Non- significant
March 2011 Non- significant Centred seedling volume (p = 0.01) Tall heath (p = 0.03) (14.9%)
Centred seedling volume (p = 0.03) (9.2%)
June 2011 Non- significant Tall heath (p = 0.02) Hand watering (p = 0.04) Centred height (p = 0.005) (11%)
Centred seedling volume (p = 0.03) (10.9%)
Jan 2012 Non- significant Tall heath (p = 0.02) Hand watering (p = 0.04) Centred height (p = 0.005) (11%)
Centred seedling volume (p = 0.02) (10%)
March 2012 Non- significant Tall heath (p = 0.01) Centred seedling volume (p = 0.005) (7 %)
Centred seedling volume (p = 0.01) (13%)
July 2012 Non- significant Centred seedling volume (p = 0.009) (13.5%)
Centred seedling volume (p=0.01) (6.4%)
Appendices
160
Monitoring date Planting substrate / planting season treatment combinations
Potting / Early planting Potting / Late planting Sand / Early planting
Initial Non- significant Monthly watering (p= 0.03) (LR/R2 = 0.01)
Non- significant
November 2010 Non- significant Non- significant Non- significant
June 2011 Non- significant Non- significant Low heath × monthly watering (d.f. 12; p = 0.01) Tall heath (d.f.12; p = 0.001) Monthly water (d.f.12; p = 0.009) (LR/R2=0.26)
January 2012 Tall heath (d.f. 15; p = 0.007) (LR/R2 = 0.26)
Tall heath (d.f. 15; p = 0.02) (LR/R2 = 0.12)
Tall heath (d.f. 15; p = 0.0002) (LR/R2=0.46)
March 2012 Tall heath (d.f. 15; p = 0.004) (LR/R2 = 0.22)
Tall heath × monthly watering (d.f. 12; p = 0.02) Tall habitat (d.f. 12; p = 0.002) (LR/R2 = 0.12)
Tall heath (d.f. 14; p = 0.0017) Hand watering (d.f. 14; p = 0.05) (LR/R2=0.32)
July 2012 Tall heath (d.f. 15; p = 0.004) (LR/R2 = 0.32)
Tall heath (d.f. 15; p = 0.004) (LR/R 2= 0.24)
Tall heath (d.f. 14; p = 0.0023) (LR/R2 = 0.42)
Appendix 5 –The optimal linear mixed models with seedling volume (log10) as a response variable and microhabitat type (bare, low, tall heath) and
watering frequency (monthly and weekly) as explanatory variables. Separate models were created for the three planting substrate/ planting season
treatment combinations and each month was analysed in a separate model. All models initially included all explanatory variables and their interactions
and then deleted in a step-wise manner until the optimal model was found (lowest AIC). The levels ‘weekly’ and ‘bare’ were used as references to
compare the plot factors ‘watering’ and ‘habitat’. The variables included were significant in the optimal model (p<0.05). The models with no significant
variables are also indicated. A likelihood ratio R2 in brackets was used to analyse goodness-of-fit for each optimal model. These pseudo- R squareds are
used to compare full model with optimal model and cannot be compared between models (UCLA: Statistical Consulting Group, 2011).
Appendices
161
Appendix 6 was submitted as supplementary material as part of Chapter 5
Appendix 6 – Comparison of osmotic potential ± SE between seedlings (n = 5) and con-specific
mature plants (n = 5) collected from the same site. Sites were classed based on their aridity
indexes (AI). Species were further categorised based on their average height as mature
vegetation (low 0–1 metre, mid 1–2.5 m and high > 2.5 m) and rooting strategy (adapted from
Crombie et al., 1988; Dodd et al., 1984; Grieve & Hellmuth, 1970; Groom et al., 2000; Groom,
2004; Lamont & Bergl, 1991; Pate & Bell, 1999). Bold values indicate significant differences
between the osmotic potential of mature and seedling con-specifics during the same sampling
date (p<0.05).
Site Species Height class
Rooting Strategy
Osmotic Potential (MPa)
Adult
(Nov 12) Seedling (Nov 12)
Adult (April 13)
Seedling (April 13)
Wandoo Banksia sessilis H Sinker roots -4.47 ± 0.63 -2.99 ± 0.40 -3.10 ± 0.42 -2.38 ± 0.23
(AI= 12) Grevillea synapheae L Unknown -3.25 ± 0.44 -2.82 ± 0.25 -2.51 ± 0.28 -1.74 ± 0.16
Eneabba Acacia blakelyi M Unknown -2.38 ± 0.31 -1.38 ± 0.13 -1.91 ± 0.19 -1.40 ± 0.15
(13) Adenanthos cygnorum
M ~200 cm -2.18 ± 0.21 -2.08 ± 0.15 -1.72 ± 0.20 -1.57 ± 0.22
Xylomelum angustifolium
H Sinker roots -2.53 ± 0.43 -2.58 ± 0.33 -2.16 ± 0.10 -2.01 ± 0.20
Motague Adenanthos cygnorum
M ~200 cm -1.84 ± 0.22 -2.35 ± 0.17 -1.56 ± 0.19 -1.50 ± 0.15
(13) Banksia nobilis subsp. nobilis
M Unknown -3.17 ± 0.43 -2.30 ± 0.59 NA NA
Banksia proteoides M Unknown -3.64 ± 0.21 -2.89 ± 0.23 -2.54 ± 0.24 -2.97 ± 0.48
Eucalyptus accedens H Sinker roots -3.65 ± 0.21 -3.15± 0.44 -2.40 ± 0.43 -1.82 ± 0.10
Gastrolobium trilobum
M Unknown -3.64 ± 0.45 -2.19 ± 0.16 -4.03 ± 0.43 -1.92 ± 0.20
Boundain Acacia stenoptera L Unknown -2.63 ± 1.09 -1.52 ± 0.08 -2.23 ± 0.23 -1.81 ± 0.19
(15) Daviesia cardiophylla L Unknown NA NA -2.44 ± 0.40 -1.76 ± 0.15
Muja Acacia pulchella L 10–42 cm -2.44 ± 0.17 -2.49 ± 0.25 -2.60 ± 0.31 -3.75 ± 0.74
(29) Corymbia calophylla H Sinker roots -2.76 ± 0.27 -2.27 ± 0.38 NA NA
Hakea ungulata M Unknown -1.94 ± 0.17 -2.92 ± 0.28 -2.24 ± 0.17 -2.60 ± 0.42
Tone Acacia pulchella L 10–42 cm -4.33 ± 0.79 -2.10 ± 0.37 -2.30 ± 0.20 -1.64 ± 0.21
(39) Banksia sessilis H Sinker roots -2.94 ± 0.22 -2.42 ± 0.19 -2.33 ± 0.27 -2.14 ± 0.28
Bossiaea eriocarpa L Shallow -2.78 ± 0.45 -3.06 ± 0.70 -2.55 ± 0.66 -1.71 ± 0.22
Phyllantus calycinus L Medium/
lateral roots -4.64 ± 0.59 -3.29 ± 0.90 -1.84 ± 0.21 -1.16 ± 0.06
Trymalium ledifolium L Unknown -4.08 ± 0.22 -2.75 ± 0.49 -2.38 ± 0.11 -2.73 ± 0.23
Northcliffe Acacia pentadenia M Unknown -2.15 ± 0.23 -2.11 ± 0.50 -1.66 ± 0.09 -1.23 ± 0.15
(44) Boronia gracilipes L Unknown -2.10 ± 0.23 -2.05 ± 0.37 -2.18 ± 0.23 -1.53 ± 0.22
Corymbia calophylla H Sinker roots
(6–15 m) -2.05 ± 0.24 -1.81 ± 0.13 -1.74 ± 0.13 -1.30 ± 0.12
Taxandria parvicseps M Unknown -1.27 ± 0.09 -2.21 ± 0.38 -1.28 ± 0.12 -1.42 ± 0.22